In 2016, the candidate who won the most votes* lost the election, which has led a number of consternated people to calculate the percentage of American voters who voted for her rival. There are a couple of different ways to do this math:
- The number of voters who voted for each candidate divided by the total number of voters who voted in the 2016 election.
- The number of voters who voted for each candidate divided by the total number of registered voters.
- The number of voters who voted for each candidate divided by the total number of eligible* voting age citizens in the United States.
Each of these measures a slightly different thing. 1 measures the strict election results without any comparison for voter turnout. 2 & 3 measure voter turnout in slightly different ways, 2 representing people who care enough to actually register. Method 3 provides the most contrast and insight: 25% of eligible Americans voted for the candidate who won the 2016 election.
That’s a startling number that throws a dark shadow over our elections. It means that 75% of Americans are not represented by the results of that election or the person currently holding the office of president.
This got me to thinking: How does this compare to previous elections? How systemic is this problem of voter turnout? Which president got the most (highest percentage of) Americans to vote for them?
Election | Voting Age Population (VAP) | Popular Votes | Popular votes as % of VAP |
2016 | 250,056,000 | 62,979,636 | 25.2% |
2012 | 235,248,000 | 65,446,032 | 27.8% |
2008 | 229,945,000 | 69,456,000 | 30.2% |
2004 | 219,553,000 | 62,028,285 | 28.3% |
2000 | 209,787,000 | 50,456,002 | 24.1% |
1996 | 196,789,000 | 47,402,357 | 24.1% |
1992 | 189,493,000 | 44,909,889 | 23.7% |
1988 | 181,956,000 | 48,886,097 | 26.9% |
1984 | 173,995,000 | 54,455,075 | 31.3% |
1980 | 163,945,000 | 43,642,639 | 26.6% |
1976 | 152,308,000 | 40,825,839 | 26.8% |
1972 | 140,777,000 | 46,740,323 | 33.2% |
1968 | 120,285,000 | 31,710,470 | 26.4% |
1964 | 114,090,000 | 42,825,463 | 37.5% |
1960 | 109,672,000 | 34,227,096 | 31.2% |
1956 | 104,515,000 | 35,581,003 | 34.0% |
1952 | 99,929,000 | 33,778,963 | 33.8% |
1948 | 95,573,000 | 24,105,695 | 25.2% |
1944 | 85,654,000 | 25,602,505 | 29.9% |
1940 | 84,728,000 | 27,243,466 | 32.2% |
1936 | 80,174,000 | 27,476,673 | 34.3% |
1932 | 75,768,000 | 22,821,857 | 30.1% |
1928 | 21,392,190 | ||
1924 | 15,725,016 | ||
1920 | 16,147,249 | ||
1916 | 9,129,606 | ||
1912 | 6,293,454 | ||
1908 | 7,678,908 | ||
1904 | 7,623,486 | ||
1900 | 7,207,923 | ||
1896 | 7,104,779 | ||
1892 | 5,556,918 | ||
1888 | 5,439,853 | ||
1884 | 4,874,986 | ||
1880 | 4,454,416 | ||
1876 | 4,036,298 | ||
1872 | 3,597,132 | ||
1868 | 3,012,833 | ||
1864 | 2,213,665 | ||
1860 | 1,866,452 | ||
1856 | 1,838,169 | ||
1852 | 1,601,274 | ||
1848 | 1,360,099 | ||
1844 | 1,337,243 | ||
1840 | 1,275,016 | ||
1836 | 762,678 | ||
1832 | 687,502 | ||
1828 | 647,286 | ||
1824 | 108,740 | ||
1820 | |||
1816 | |||
1812 | |||
1808 | |||
1804 | |||
1800 | |||
1796 | |||
1792 | |||
1789 |
The results are:
- 2016 was a (slightly below) average election.
- The problem is incredibly systemic and long term.
- LBJ got 37.5% of Americans to vote for him in 1964. (And here I thought it'd be FDR, but at his best in 1936 he only managed to get 34.3% of Americans to vote for him.)
- GHW Bush only got 23.7% of Americans to vote for him.
And the answer is:
- It only takes 23% of America to elect a president.
This is important because it means that in any given presidential election, only 1/4 of America is being represented by the winner. Tyranny of the minority.
But this result makes sense because of the electoral college. Because of the winner-take-all system of allocating electoral college votes used by 48 of the States (as of this writing), 50% of any given state's voters are ignored by the election. Don't believe me? Try voting for a 3rd party candidate and then see how many of your state's electoral college votes are allocated to that candidate. If that 50% of voters whose vote literally doesn't count chooses to not vote rather than to vote for a losing candidate, we wind up with 50% voter participation. The remaining 50% of voters who do participate split 50/50 (ish) based on party affiliation. So, if this hypothesis about the electoral college is correct, we should see somewhere in the neighborhood of 50% voter participation and an average of 50/50 splits between the major two parties. The actual numbers in any given election vary but hover around 55.7% voter participation and 48/47% splits between D/R parties, which is pretty close to my predictions.
(Parenthetically in parentheses, the D/R parties both average below 50% because there are 3rd parties in the mix pulling single digit percentages of the popular vote every couple of elections. And these numbers are for the popular vote, not electoral college electors.)
It's also worth noting that this winner-take-all low-voter-turnout system caused by the electoral college also encourages a bipolar party system that severely limits the ability for a 3rd party candidate to make significant progress in a presidential election. Simply put: because of the electoral college, any 3rd party has to completely usurp one of the existing major parties in order to have any shot at winning even a single state. Utah is a good example of this: in 2016, 21.54% (almost 1/4) of Utah voted for Evan McMullin - but can you guess how many electoral college votes he got? That's right, ZERO.
It's also very important to recognize that this isn't just an electoral college problem. Poor voter participation and the two-party system are further reinforced by Article II of the United States Constitution and the 12th Amendment. That is, 3rd party candidates (and therefore any voters who might vote for them) are hamstringed by the constitutional requirement that the winner of the election must receive greater than 50% of the votes (or the election goes to the House, but we usually forget that part). This is why the numbers in the data above all hover around 29%, because 29% of all eligible voters is just over half of the average ~55% voter participation rate.
The takeaway from this is that any given presidential candidate only needs to impress 25% of Americans to win the election. And that's a bad thing*.
* In this article I refer to Americans as voting directly for president. However, in the United States, because of the electoral college, Americans don't vote for their president. Instead, they vote for electors who then vote for the president based on rules written by the state's legislature. These electors, representatives of their state and not the people, vote for and elect the president. This is a subject of great debate, but is generally recognized as a bad idea.
* For the purposes of this article, I used VAP - Voting Age Population - because it was the statistic that was available. A more refined number, taking into account factors such as disenfranchisement, called Voting Eligible Population (VEP) exists, but the data for elections going back decades does not exist.
* The data available only took me back as far as 1932. Beyond that year, the data for the total VAP wasn't (readily) available and may not exist. My data sources include this Wikipedia article and this Encyclopaedia Britannica Online article.
* I emphatically state that it's a bad thing while not providing explicit solutions to the problem. I would recommend removing the electoral college and switching to a "most votes wins" system. (I know a lot of readers will cry "tyranny of the majority!", but I fail to see how democracy is any worse than oligarchy. Notably, all of those who would cry so are members of the ruling minority.) In a popular vote most-votes-wins electoral system, the president may still get elected by only 25% of eligible voters, but that president could be from any party and not just blue or red. And with 3rd party and independent candidates becoming more viable, voter participation will increase. It might take a few elections to happen, but in a decade or two we'd find ourselves with a congress that had no single party majority - which means no single party could arbitrarily pass or block any legislation and therefore congressional representatives would be forced to cooperate. This is common in other nations where you hear of "coalition governments", the coalition being two or more parties agreeing to work together in order to accomplish mutually beneficial goals. It is by no means a perfect solution - any electoral system is an exercise in compromise. But a compromise that involves more of America and therefore represents more of America is a better compromise than the one we're making now.
* I should probably note here that there are multiple possible voting systems that can fit under the rather basic description I have above (not just the simple plurality), including ranked choice, proportional, cardinal, and others, all of which would be better than what we have.
* I recognize that the electoral college exists for a reason. But the form of government the founders intended when they wrote the Constitution, namely a tight federation of independent states as opposed to the loose confederation that existed under the Articles, frankly no longer exists. The United States is no longer a collection of states united by a common currency and defense (as the EU is), rather the United States is a singular state, and our electoral system needs revised to reflect that. (Though I'll happily entertain educated and rigorous debate that we should abolish modern federalism and return to that original intent. I don't think that's practical but I'd love to examine it.)
Category: