What effect did the Electoral College really have in 2020?
With 2020 Census data published, we can take a deeper look
Much noise was made during election season by left-leaning partisans about the Electoral College, the somewhat byzantine system in the US by which Presidential votes are counted to elect a President. Frustrated Democrats shared infographics about how many “Californians” are worth one “Wyomingite” and gnashed their teeth over what they called an un-democratic system.
The Republican rebuttal, interestingly, generally conceded the math and argued instead that this disproportionate representation is a feature, not a bug. Now personally, I suspect “principled” arguments about “tyranny of the majority” or “undemocratic” systems are really just post-hoc rationalizations for whichever position maximizes your favored party’s chances of winning. But beyond partisan bickering, there’s the question of just how big (or small) a difference the Electoral College actually makes in determining the winner of elections.
Is it true that the EC overcounts Republicans while undercounting Democrats? Do those infographics from angry Democrats about Californian-to-Wyomingite exchange rates have it right? With the US Census Bureau publishing state-level residence data for the 2020 Census, we can compare population numbers to 2020 election results and see for ourselves. And in true click-bait fashion, the answers may surprise you…
Caveats
Reminder that both election results and population counts change, so this analysis should be seen as specific to 2020, though I think looking at a recent snapshot should be directionally accurate for near-future elections. Also keep in mind that the size of an effect should not be conflated with its impact. If elections are very close, then a modest sized effect can be enough to swing a margin and flip an outcome, which would be a huge impact. Finally, your obligatory reminder to enjoy data analysis responsibly. Do not drink and dis-aggregate.
Methodology
Note that all backup data and calculations can be viewed in Google Sheets here.
What makes a state “undercounted” or “overcounted”? Since states (mostly) use winner-take-all systems to award their Electoral Votes based on statewide popular vote, we can compare the % of total US population a state represents vs. its % of total Electoral Votes allotted. For example, California’s population of 39.5M people represents 12% of the total US population of 331.4M, but its 55 electoral votes are only about 10% of the total 538. It turns out the Golden State is indeed undercounted, but by how much? To calculate, we divide its EC share by its Population Share: 10.22% / 11.93% = 0.86x. The smaller the ratio is than 1.0x, the stronger the undercount.
(Note that these numbers are slightly different due to rounding; look at the Google Sheets for exact figures).
Conversely, Wyoming’s 577K population comprises 0.2% of the total US, but its 3 electoral votes represent 0.6% of the total 538. 0.56% divided by 0.17% equals 3.2x. The greater the ratio is than 1.0, the stronger the overcount.
Results
There are 19 states that are “under-counted” by the EC. These states comprise 74% of the population but 67% of Electoral Votes. The undercount ratios vary from 0.80x to just under 1.00x. We can call this bucket the “Moderate Undercounted”. The remaining 32 states (plus DC) can be split into two equal groups of 16: the Moderate Overcounted (1.00x to 1.26x) and the Strong Overcounted (1.34x to 3.20x).
Let’s see how the parties did across these three groups.
Democrats indeed won the (Moderately) Undercounted states 13 to 6, virtually 2:1, picking up 66% of the Electoral Votes in this bucket. Conversely, Republicans won the Moderately Overcounted states 11 to 5, a similar 2:1 margin picking up 67% of these Electoral Votes.
But surprisingly, when it came to the 16 Strongly Overcounted states, Democrats and Republicans split these almost exactly 50-50 and Democrats in fact picked up 51% of the Electoral Votes to the Republicans’ 49%.
These states include: Idaho (R), New Mexico (D), Nebraska (split 80-20 to the Republicans), Hawaii (D), Montana (R), West Virginia (R), New Hampshire (D), Maine (split 75-25 to the Democrats), Delaware (D), South Dakota (R), Rhode Island (D), North Dakota (R), Alaska (R), DC (D), Vermont (D), and Wyoming (R).
The states that are given the most outsized influence in the Electoral College help Democrats and Republicans identically.
It’s really just the middle third that skews Republican. If we combine all the Overcounted states, Moderate and Strongly, we get a modest 60-40 Republican lean.
Moreover, it’s not just an artifact of grouping the states by thirds that we see relatively even skews. If we look at the Top 3 Undercounted States and the Top 3 Overcounted States, we actually get the exact opposite trend from the common accusations.
The Most Undercounted:
And the Most Overcounted:
If we looked only at the Worst Offenders, we’d expect to see Republican Texans complaining about Democratic Vermonters as much as Californians kvetch about Wyomingites.
Comparing the Parties
We’ve been looking at three distinct groups of states to see the partisan lean of under vs. over-counted states. But when you add it all up, were the Democrats truly handicapped by Electoral College apportionment in 2020, or by the Electoral College in general?
Let’s see.
Democrats won 26 states (and DC), representing 57% of the population, and received… 57% of total Electoral College Votes. By contrast, they only won 51% of the total popular vote.
In 2020, Democrats had no handicap due to Electoral College apportionment, and in fact outperformed their popular vote performance in the Electoral College.
The Upshot
The Electoral College and its apportionment in 2020 indeed weighted states disproportionately relative to their populations. However, the most disproportionately boosted small states benefitted Democrats and Republicans identically. And overall, the EC results were exactly representative of Blue/Red state populations.
Of course, I don’t expect this to stop partisans from making their impassioned arguments in favor or opposed to the Electoral College or its apportionments. But maybe the next time you see Democrats complaining about those Wyomingites or Dakotans, you can remind them not to let Vermonters or Rhode Islanders off the hook, either. And when Republicans wax philosophic on countering the Tyranny of the Majority, you can point out if that’s truly the goal, the EC isn’t all that effective.
I can’t promise you’ll make many friends, but hey, at least the Census tables will never block you on Instagram.
—
Nihaar Sinha is a screenwriter and Founder/CEO of Indo-SoCal lifestyle brand Maiya. He lives in Los Angeles, CA.
The misapportion of EC votes is not nearly as big of a sin as the winner-take-all part of the electoral college. Indeed Trump still would have won 2016 if CA had 12% of EC votes, WY 0.6% etc.
This analysis misses the forest for the trees. Yes, this is a flaw in the EC that ends up being a wash in partisan advantage. But the main flaw that causes divergence from the popular vote is not.