Winner-take-all voting underpins U.S. system

Credit: India Price/Online Editor Credit: India Price/Online Editor
Editorials featured in the Forum section are solely the opinions of their individual authors.

Editor’s note: The Pragmatist’s Manifesto is a weekly column that encourages civil discussion about politics in the U.S. Find previous installments @

Since our current election has hit … let’s just say a low moral point, I am going to do my best to discuss anything else in the world of politics. Last week we discussed the Electoral College, so this time I want to dive further into the voting system that the United States has had for over 150 years. We’ll discuss its precedents, alternative systems, and why the one we have works the best for our country. I’m talking about that much criticized system: first-past-the-post, or winner-take-all voting.

Winner-take-all systems are just that the person who receives the most votes in an area wins the seat, or, in case of the presidential race, all of that state’s electoral votes. We discussed this with regards to the Electoral College, but by itself, winner-take-all races are prevalent all throughout our nation at all levels of government. When several candidates run for one office, however, it raises the possibility that one candidate will win with a plurality — less than 50 percent of the vote.

In our modern times this has led to primaries and runoffs, in which only the top two finishers in the primary face off in the general election. This system works, but usually primaries do not get as high of a turnout. Typically, only more involved and passionate voters take part.

We like to think of our country as a hallmark of democratic values, but things did not use to be so democratic. In terms of the presidency, electors were selected in a variety of ways. There is nothing in the Constitution that mandates using a winner-take-all system. The nonprofit organization FairVote describes how selection by state legislature was the preferred method for several decades in our early history. This was another system that the founders established to prevent mob rule. After the “corrupt bargain” in the election of 1824, more states shifted towards popular election of electors.

This began with certain states changing their systems for mainly partisan reasons to enforce the two-party system. Once Democratic-leaning states changed their selection to winner-take-all system, Republican- (or at the time Whig-) leaning states would do the same.

More recently, in political races including the presidency, primaries are held to find the top two candidates so people can choose between two, preferably moderate, sides. This is not always the case, looking at our current election.

However, alternative voting strategies exist, and there are a wide variety of them. We will discuss two: Approval and Instant-Runoff Voting.

Approval Voting is simple. Any number of candidates run for an office, and people can vote for as many as they want. Whoever wins the most votes wins. Simple.

Although it seems like a better system, there are several practical reasons why this may not work for our country. These two systems tend to lower amounts of negative campaigning, but as we have seen, campaigns get heated. If there’s one thing Bernie Sanders taught us, it’s that candidates running a positive campaign will sometimes turn sour when the going gets tough.
If multiple candidates ran from each major party for any office on any level, Approval Voting could run the risk of “bullet voting,” or only voting for the candidate you care most about. It is realistic to think that candidates would encourage this in the heat of an election.

Each vote on these ballots is equally weighted. Many voters would not want to equally support candidates in practice; this leads itself again to plurality voting as we have it. Approval Voting also does not eliminate the idea of spoiler candidates and would make us drop the idea that 50 percent plus one is a winning coalition, as many candidates might get more than a majority.
The next system is Instant-Runoff Voting, which is a bit more complicated. It requires voters to rank all of the candidates from favorite to least favorite. Whichever candidate is last after tallying everyone’s top favorite gets eliminated, and those ballots move to their second favorite. This process continues until a candidate gets a majority of votes.

Although this system can work, it’s not a practical solution for logistical reasons. For one, ballot counting. If there’s one thing that elections in the past have taught us about ballot accuracy (2000), it’s that ballots can be inaccurate. In addition, voting in the United States is drastically underfunded. Realistically, counting these ballots quickly and then having to find a way to move millions of ballots to a second or third choice would be extremely complicated and time consuming. Until Congress allocates more money for changing the way we vote in elections, adjusting to this system would be impossible.

Another key issue with ranked voting is Arrow’s Impossibility Theorem. Yeah, you just got some serious math thrown in here, so get ready. Named for Nobel-Prize winning economist Kenneth Arrow, the Theorem postulates that no ranked voting system could be designed that satisfies three “fairness criteria.” it explains these as being:

That if all voters in a group prefer candidate A over candidate B, the group then prefers A over B.
If a candidate is eliminated, then the choice of candidates should depend on the voter’s preferences among the remaining candidates.

There is no single voter able to determine the outcome for a group.

I won’t get into the extended details of the proof for the Theorem, but you are welcome to look into it if you, like me, like to geek out on election science.

Looking at these systems brings up the question: How is the plurality voting we have better than this? Sure there might be faults, but it is an improvement.

As we have discussed at great length, the system established by our founders lent itself to a two-party balance. These two parties, although ideologically different, can be effective if they endeavor to work together. These alternative voting systems do not yield to a healthy two-party balance.

The French sociologist Maurice Duverger hypothesized Duverger’s Law in the mid 20th century. This law states that plurality voting such as first-past-the-post tends to encourage two-party systems. Some political scientists since then have theorized the inverse; that the two-party system may in fact favor first-past-the-post voting.

In a situation in which candidates run for an office with something other than plurality voting, there are a host of other threats that emerge, including bullet voting and strategically designating candidates as lower preferences to ensure they do not get support if a voter’s first preference does not win. Winner-take-all elections bring more voters under the two-party umbrella.

And again, there are too many logistical and financial obstacles to implementing one of the other voting systems.

We also have to be realistic about how people vote. Many times, as this year has shown us, people do not vote rationally. By using alternative voting, we would run the risk of having someone elected who does not have real popular support.

I still hope that after Nov. 8, our parties can begin to shift back to a degree of sanity and compromise. Winner-take-all is by no means a perfect system for voting. But in the United States, its benefits for a two-party system cannot be ignored.