How Trump Broke the Electoral College, and how Democrats can Unbreak it
After the 2012 election, many believed it would be a permanent struggle for Republicans to win the Electoral College. Barack Obama had just crusied to a relatively comfortable re-election. Not only did he win the national popular vote by almost 5 million votes, but the Electoral College was slanted in his favor, with the margin in the tipping-point state, Colorado, sitting at 5.4% — about 1.5% higher than his popular vote margin. Furthermore, Obama performed very well with non-white voters, and the non-white share of the electorate increases every cycle.
Trump made quick work of this theory by doing what right-wing populists have done throughout the Western World in recent years: winning over working-class white voters that were previously sympathetic to left-of-center parties.
Even though the United States is much more diverse than say, the United Kingdom, this was a viable strategy for Trump, because America’s electoral instututions allow working-class white voters to punch above their electoral weight. This is particularly true of the Senate, and will be for a long time, but it has also been true of the Electoral College in recent cycles. Because of this, Trump’s massive 2016 gains in rural areas propelled him to the Presidency even as he lost the popular vote by 3 million votes.
In 2020, the Electoral College nearly allowed Trump to win re-election, despite losing the popular vote by 7 million votes. Doing 0.7% better in Arizona, Georgia and Wisconsin would have given Trump a second term.
How Trump Broke the Electoral College
The structure of the Senate and Electoral College both became significant obstacles for Democrats in the Trump era, but for different reasons. The Senate over-empowers smaller states, most of which are solidly Republican, with a few exceptions (New Hampshire, Maine, New Mexico). This has been the case for awhile, but what changed during the Trump era is that increased polarization made it more difficult for Senate candidates, even incumbents, to significantly outperform a State’s baseline partisan lean, especially in a Presidential election. This is a big part of why Heidi Heitkamp, Claire McCaskill and Joe Donnelly are no longer Senators, why a popular Governor in Steve Bullock didn’t come close to winning in Montana, and why Joe Manchin and Jon Tester will be underdogs in 2024.
The Electoral College, on the other hand, does not substantially over-empower the likes of North Dakota, Montana or West Virginia — they don’t have many electoral votes. Due to the winner-take-all mechanism, the Electoral College over-empowers medium-to-large states that happen to be close, while under-empowering medium-to-large states that are extremely uncompetitive.
Florida, Pennsylvania and Ohio have historically fit the role of “large swing state”, and as such have been the biggest Electoral College bellwethers. Trump flipped all three of them in 2016, and held on to two of them in 2020. Additionally, Trump’s weakness in large Sun Belt metros turned Texas from a Safely Republican state into a competitive state, but he still won it twice.
The flip side is that the Electoral College does not reward candidates for racking up huge margins in large, uncompetitive states. Democrats do this in a number of states, but the most impactful by far is California. Clinton won California by 7% more than Obama, and Biden won it by 6% more than Obama, which translates to over a million additional votes in margin at 2020 turnout levels. All those extra votes in California did nothing to help Clinton or Biden, while Trump’s improvements in Ohio, Pennsylvania and Florida, as well as medium-sized swing states like Michigan and Wisconsin, almost made him a two-term President with zero popular vote wins.
But wait, it gets worse!
Just when you thought things couldn’t get worse, Electoral Reapportionment is going to be a net transfer of Electoral Votes from Biden-won states to Trump-won states, thanks to Solidly Democratic states like New York, California and Illinois losing population to Texas and Florida. Biden won 306 EVs, but he would’ve only won 301 or 302 under the 2024 Electoral Map*. Because of this, there are plausible maps that result in a Democratic win under 2020 EV apportionment, but a Republican win under 2024 EV apportionment. For Example:
* We won’t know the exact map until the Census results are finalized, the uncertainty surrounds whether New York will lose 1 or 2 Electoral Votes
How Democrats can Elimate the Electoral College Bias
There are two basic pathways by which Democrats can eliminate the Electoral College bias, or at least substantially reduce it. This is illustrated nicely by this diagram from Nate Silver:
Scenario 1: Reverse Trump-Era Trends
The first path is to partially reverse the trends of the past 8 years. By making up lost ground in Pennsylvania, Michigan, Wisconsin and Nevada, Democrats could eliminate the EC bias, and win comfortably even if they were to backslide in Georgia, Arizona and Texas:
Under this scenario, Democrats might do slightly worse in the popular vote than Biden, and would win fewer EVs. But their Electoral College victory would actually be much more comfortable, because the tipping point state would be Nevada with a margin of almost 5%, compared to the 2020 tipping point of Wisconsin at 0.64%.
It’s an open question how realistic this scenario is. It’s possible that white working-class voters are gone, and there’s not much hope to hold onto the Midwest long-term, particularly Wisconsin. But I think it’s at least plausible that a Joe Biden presidency that spends trillions helping the lower half of the income distribution, doesn’t raise taxes on the middle class, and generally avoids “culture war” topics could create *some* educational depolorization, which is the key to this scenario.
Scenario 2: Turbocharge Trump-Era Trends
The second path is to go in the exact opposite direction: turbocharge the Trump-era trends by continuing to gain in Arizona, Georgia, North Carolina, and most importantly, Texas.
Under this scenario, Texas would become the tipping point state, and a huge weapon for the Democrats, allowing them to crater in the Midwest and lose Nevada while still winning the Electoral College, and potentially wiping out most of the Electoral College bias.
But there’s a danger to taking this pathway: as Nate’s diagram illustrates, it requires larger swings than the “reverse Trump-era trends” path, and it’s possible to get stuck in an in-between point where the Electoral College bias is *even worse* than it was in 2020. Here’s an example of that:
Scenario 3: Trump-era Trends Continue, but not enough to Flip Texas
This scenario follows the same direction as Scenario 2: Dems improve in the suburbs of large Sun Belt metros, but do worse with working-class voters. The only difference is that the swings are smaller, most crucially in Texas, narrowly keeping it in the Republican column and allowing the GOP to reach exactly 270 Electoral Votes while getting blown out in the popular vote. This is bottom of the curve in Nate’s diagram, and it seems just as plausible as the other 2 Scenarios for 2024.
What about Florida?
One thing you may have noticed about the above scenarios is that they don’t discuss the always-critical state of Florida. The basic reason for this is that it’s a weird state, and I’m unsure how it fits into any of these scenarios.
Florida could play a similar to role to Texas: if Democrats can flip it, they could win the Electoral College while losing most of the Midwest. And if they get Florida roughly in line with the popular vote, it could eliminate most of the EC bias.
But I think that’s a tough ask. Florida has been at least 2.5 points to the right of the popular vote in five straight presidential elections, and it took a big step right in 2020. There are some great trends in Florida for Republicans: there’s a constant stream of retirees moving there, and Trump made massive gains with Florida Hispanics, most notably in Miami-Dade county, but also in other areas such as Osceola County in the Orlando area.
Not all the trends in Florida are good for Republicans: Biden flipped Duval County (Jacksonville) and Seminole County (northern Orlando suburbs), and improved in the panhandle. But while Florida could be a wildcard in any of the above scenarios, there’s a good chance it stays put, with counterveiling trends cancelling each other out, and continues to be elude Democrats by narrow margins.
Unless Biden’s Extremely Popular, 2024 will probably be another Grinder
The upshot of all this is that there’s a pretty good chance that the 2024 Presidential Election will be another close one, and that the tipping point state will be to the right of the popular vote once again. To be sure, there is much uncertainty, including the question as to whether Biden and/or Trump will run again. But the possibility of another electoral college / popular vote split is likely to loom large until Democrats manage to either win back a large chunk of Obama-Trump voters, or culminate the Sun Belt project by bringing Texas in line with the national vote.
