Even Without The Electoral College, The Smaller Places Matter
In the United States, there have been five Presidential elections in the last 20 years. Of those five, two have been quite peculiar: the winner of the election — that is, the one who captured the majority of votes — was not declared the winner of the election — that is, the one who captured the Presidency. Such is the nature of the Electoral College, the idiosyncratic system by which we choose our President. As one may expect, this process is controversial with many defenders and detractors, even to the point of some states attempting to de facto abolish the system entirely.
To those wishing to end the Electoral College, the argument is simple: the Electoral College undermines the core idea that in an election, every vote counts. In a CNN town hall event in March, Massachusetts Senator and Presidential candidate Elizabeth Warren expounded on this idea:
Come a general election, Presidential candidates don’t come to places like Mississippi. They also don’t come to places like California and Massachusetts because we’re not the battleground states. My view is that every vote matters, and the way we can make that happen is that we can have national voting and that means get rid of the Electoral College — and every vote counts.
Indeed, given the allocation of Electoral College votes, it makes little sense for Presidential candidates to visit non-battleground states. In the vast majority of them, the winner of the state receives all of its Electoral votes. For the millions of people who voted Republican in California and New York, Democratic in Texas and Florida, even the hundreds of thousands of voters who voted for a losing candidate in states like Tennessee, Oregon, Maryland, and Louisiana, their votes are not reflected in the final Electoral College votes. Unless there is a chance to outright win the state (i.e. a battleground state), the additional support a candidate may gain visiting a state does not affect the result; in effect those votes do not count.
To the defenders of the Electoral College, the argument is equally simple: without the Electoral College only big states and cities matter. Rural communities would be rendered irrelevant in the Presidential vote. In a video and subsequent tweets sent out by Texas Congressman Dan Crenshaw, he explained this argument:
I had some sympathy for this line of reasoning for a long time. While I didn’t really believe that the smallest states were all that prominent in the Electoral College — after all, no campaign has ever really clamored for, say, Wyoming’s electoral votes — I did think that eliminating the Electoral College would only put the focus on big states like California, Texas, and New York. At least, I thought that before seeing this thread from New York Times columnist Jamelle Bouie.
I was actually pretty surprised by this claim; for some reason, I just accepted the conventional wisdom that an election by popular vote would be disproportionately decided by the biggest states without actually checking the math. So I looked at the numbers myself to verify the numbers in the rebuttal; as it turns out, all of the numbers in the thread are correct.
At this point, I wanted to delve deeper in the numbers to completely debunk the notion that only big cities and states would matter. However, it would be time-consuming to manually collect and analyze the vote totals, especially at a county level — Texas alone has 254 of them. So I made a program to scrape the election totals by county that appeared on Wikipedia (excluding Alaska which required a slightly different process because the state doesn’t report results by county). Then using a site which helps to make cool data visualizations called ObservableHQ, I made maps for manipulating and visualizing state and county hypothetical election results given 2016 number of ballots cast. Both the ObservableHQ link of the maps and the GitHub repository of the scraper will be linked at the bottom of the page.
First, let us look at the state numbers. In the rest of the tweets of Mr. Bouie’s thread, he states that the top 10 states by population add up to about 52% of the popular vote, so if you somehow won every single vote in those states you could ignore the other ones. Contrast this with the Electoral College, where you could theoretically become president with about half the number of votes. This makes the case that big state domination is more likely under the Election College than a popular vote.
In fact, if anything, Mr. Bouie understates that case. Suppose instead of a candidate winning a state by getting every single vote, they won with 67.40% of the vote. This is equivalent to the percentage of votes won by Donald Trump in 2016 in Wyoming, a state considered to be in the heart of “Trump Country”. In order to show the results of this hypothetical, we shall introduce our two mock candidates for all of the scenarios examined: Elaine Marley and Charles L. Charles who represent the Democratic and Republican Party, respectively. For the following example, we will have Charles L. Charles be our candidate of choice to win 67.4% of the vote. To win a slim majority of total votes, Mr. Charles would need a minimum of about 40% of states, including states such as New York, Pennsylvania, Florida, Missouri, Colorado, and Indiana. In the Electoral College, this produces a result that is roughly equal to Barack Obama’s 2008 election victory over John McCain.
Now suppose we lower the winning percentage to a mere 61.73%, the percentage of votes won by Hillary Clinton in California, a Democratic stronghold during the 2016 election. Charles would need close to half the states in the county to win a slim majority.
This produces a result where the winning candidate would receive the largest percentage of Electoral College votes since 1988, the 8th largest since World War II, and the 14th largest since Reconstruction. Given the incredibly small likelihood of winning 24 states with at least 61% — not even Ronald Reagan was able to accomplish this feat in the 1984 Presidential election where he won every state except one — clearly the advantage of smaller states having influence in a national election favors the popular vote.
In fact, for either threshold, the range of possible margins of victory in the popular vote is smaller than the margin of victory in the Electoral College. No matter the distribution of votes in the states that Elaine won, the resulting margin of victory will be smaller than with the Electoral College.
Next, let us consider a scenario dubbed by Erielle Davidson, staff writer for the Federalist, as the “tyranny of the coasts”, where the person who wins the presidency is the one that gains the most votes from the coasts, completely ignoring the rest of the country. In her view, this would be the result of abolishing the Electoral College.
Senior editor at Bloomberg Derek Wallberg echoed this same sentiment in this tweet, warning the Democrats that they “can win the coasts by 30, but [they] have to win the Electoral College to win.” This implies that blowout wins in the coastal areas are not as advantageous in the Electoral College as the popular vote.
So, let’s put this “tyranny of the coasts” scenario to the test by having our hypothetical candidate Elaine Marley win all of the states on the East and West Coast by 30 percent. In order to maximize the chances of Charles E. Charles winning the Electoral College, we will have him win every other state.
(Note: there is some disagreement about whether Alaska and Hawaii are considered part of the West Coast or whether West Virginia is considered part of the East Coast. Therefore, the following analyses will consider all scenarios that include and exclude those states from the respective coasts)
Under the popular vote, Elaine has a clear advantage over Charles in being able to gain a majority. Elaine gains 46.9–47.8 million votes depending on the states included in the East and West Coast count as explained earlier, which translates to about 70% of the votes needed to gain a majority. Given that massive advantage, it would seem that the Electoral College allows for better representation of non-coastal states, right? Well, there is a small problem with that theory: despite the advantage in the popular vote, the advantage in the Electoral College is even greater.
Under the Electoral College, the only chance that Elaine has of losing the election is if we exclude Alaska, Hawaii, and West Virginia from her Electoral College count. In that case, she would just need one more state to win the Electoral College. While Elaine would need at least 21 million more votes to win given the popular vote system, she could theoretically win the Electoral College with a majority of Wyoming’s vote total — as little as 130,000 votes.
But, to be fair to the defenders of the Electoral College , winning the coasts by 30 points is not the same as winning every single state in them by 30 points, so, let us change the scenario. Suppose we give Elaine and Charles the same distribution of Electoral College votes as in the 2016 general election while maintaining the same 30 point advantage in the popular vote. Naturally, Elaine would still need 21 million votes to win in the popular vote. In the Electoral College, Elaine gets anywhere from 178–182 Electoral votes, approximately 66–67 percent of the necessary votes for a majority, virtually the same share of votes needed for a majority as the popular vote (technically a bit less but a single state overcomes that small difference).
And so, we see from these maps, the “tyranny of the coasts” is no tyranny at all. At best, the Electoral College represents the non-coastal states as well as the popular vote. In fact, in some instances, it does worse.
But really, the “tyranny of the coasts” concept is a specific application of a more general defense of the Electoral College: greater regional representation. The common wisdom is that Electoral College allows the regions in the United States to have more equal representation during an election.
One problem with this claim is each region has roughly the same proportion of total votes for the Electoral College and the popular vote. Here, the regions of the United States are defined in the same manner as the Census Bureau. Below is the breakdown of the Electoral College and popular vote percentage of each region and division of the U.S.
As the table shows, switching from the Electoral College to the popular vote does not significantly change the percentage of the total votes from each region. In short, the Electoral College and the popular vote provide the same regional representation by percentage.
The other problem with this claim is that due to the winner-take-all nature of Electoral vote allocation, some regions actually have their power limited in a way that does not happen in a popular vote. Consider the West region of the United States. Of the states in the regions, only Colorado and Nevada are considered perennial swing states since 2000, a total of 15 electoral votes. Contrast this with the Midwest or the South that have multiple battleground states with more Electoral votes at stake.
The result of this is that presidential candidates will spend a lot more time in the other battleground areas. This makes sense because, as explained earlier, it’s not really helpful to spend time in a state that is not projected to actually be a close race in the Electoral College. This was borne out in 2016 presidential campaign events, where the bulk of events were done in battleground states such as Florida, Pennsylvania, and Ohio. As a result, the Western states lose influence; the concerns of the people in the region are neglected.
Now consider those same Western states under a popular vote system. Unlike the Electoral College, there is incentive for politicians to visit these areas. While the winner of the state may be the same as the Electoral College, a candidate can pick up hundreds of thousands to potentially millions of votes towards their national total under the popular. It is not hard to imagine a Republican candidate doing events in New Mexico and Oregon or a Democratic candidate doing events in Utah and Idaho to pick up some votes. This way, the concerns of that region are better heard; they have more power to influence national candidates.
As with the “tyranny of the coasts” scenario, the popular vote provides equal regional representation to — if not greater than — the Electoral College.
Another defense of the Electoral College is that only big cities matter. In order to properly examine this claim, we will analyze hypothetical elections at the county level as opposed to the state level. At this level, the minimum number of counties needed to win a slim majority of votes is 156. While this does represent only about 5% of all counties in the United States, there are two things to keep in mind before declaring the Electoral College superior in representing the entire country. The first is that this scenario is also enough to win a slim majority in the Electoral College, 282–256. In fact, a slimmer margin of victory is possible with the popular vote result as opposed to the one for the Electoral College.
The other is that the top 156 counties in the United States represent every county larger than Genesee County — the county of Flint, Michigan — which cast about 196,000 votes in 2016. Examples of counties that are larger than this one include Bernalillo County, New Mexico, Lucas County, Ohio, and Douglas County, Nebraska. When people argue that big cities would dominate without the Electoral College, they mention New York City, Los Angeles, and Chicago, not so much Albuquerque, Toledo and Omaha. And yet, even if a candidate got every single vote in New York City and Los Angeles, they would still need votes from places like Des Moines and Tulsa.
As was done with the state data, below are the results with the voting share of the winning candidate at Donald Trump’s share of Wyoming and Hillary Clinton’s share of California. The thresholds for a majority of votes are 471 and 671 counties, which includes every county larger than 57,931 and 37,425 votes, respectively. While the popular vote does not translate into as strong of an Electoral College victory as the state data, the results are similar to the state data: a larger margin of victory in the Electoral College than the popular vote.
At a county level, it is easier to see whether the rural area is actually represented better in the Electoral College than the popular vote. Here, “rural area” is defined as a county with under 50,000 votes. As with the previous claims, this one is also inaccurate. To understand this, recall that in the Electoral College, cities only affect the race in their current state, as opposed to the national race under the popular vote. Now, consider states such as Texas and Illinois where there are millions of rural votes available — or even Oklahoma with hundreds of thousands of votes— and have had mostly blowout election results for one party for the past 30 years.
Which would be better for those voters: the Electoral College, where those votes are confined within their states, often in contests that are not in doubt, or the popular vote, where voters from different states can link together and form a coalition among the 32 million rural voters? This is especially true for a state such as Illinois, a state in which only about 20 percent of the state vote came from rural counties and 46 percent of them voted for someone other than the candidate who won the state in 2016.
With all this in mind, let’s review some of Congressman Crenshaw’s claims:
- The Electoral College creates equal regional representation and protects the interests the less populated states
- The rural population is not ignored in the Electoral College; would be ignored in the popular vote
- Every candidate would only spend time in the 10 most populated states to get to 51%
As shown throughout, these commonly cited claims are not accurate. The popular vote does a better job of protecting regional interests, rural populations, and smaller states. Despite this, however, there may be other arguments in favor of the Electoral College. Likewise, there may be other considerations for a “one person, one vote” popular presidential voting system. However, the claim that only the big cities and states matter if the Electoral College is abolished is false.
You can verify the popular vote maps used here or create your own maps with the 2016 total vote data by going here. Maps of the Electoral College were made using 270towin.
Code used to scrape the 2016 election data can be found here. Most of the numbers used in this piece that are not directly cited were derived from a count of 2016 total votes at the state and county level, which can be found here.
