Third Party's the Charm (essay from 2017):
📅I wrote an essay last semester at UMD about a proposed system for voting that would allow people to vote for a third party without wasting their votes. Voters would indicate their first, second, and third choices for candidates and the system would evaluate the choices so that if a voter’s first choice candidate did not win, the vote would be reassigned to the second choice, and the count would be re-evaluated.
[Update 2024-11-06]
At the time I wrote the essay (2017), I didn’t know that such systems already existed and I had basically reinvented instant-runoff voting (wikipedia). One downside I didn’t anticipate was the voting method criteria that it fails to meet, although it does meet more criteria than the current first-past-the-post voting system. See this table of comparison of voting rules. Also, it looks like a instant-runoff voting system has been adopted in Maine (in 2018) and Alaska (in 2022), with some issues relating to the failed criteria (near the end of this section). It’s important to find the giants and stand on their shoulders when thinking of “new” ideas! If I did my research properly I might have been able to discuss the intricacies of different ranked voting systems instead of naively thinking of a system that solves one problem but might create another. We’ll see how things go in Maine and Alaska, and hopefully the benefits outweigh the downsides.
Also wow, I just watched a video about voting systems that explained instant-runoff voting in the exact way I recommended in the essay: “The way the voting system works could easily be explained with an animation of bar graphs representing the number of votes for each candidate. The bar belonging to the candidate with the least votes would be cut into pieces of different colors representing the second choices of the voters who chose that candidate as their first choice. Then the pieces would move to the corresponding bars for those candidates who are still in the running.”
Full Essay
ENGL101
November 30, 2017
Third Party’s the Charm
The polarization between the Democrat and Republican parties has been increasing over the past few decades. The causes of the polarization relate to the way our brains are wired, involving tribal thinking and the willingness to trust party decisions rather than doing further research. Because the polarization is so high between parties, very few people consider voting for a third party. When I tell people that I voted for a third party candidate, the most common response is either “you wasted your vote” or “I might have voted for a third party if they had a chance of winning.” Since third party candidates are not likely to win, people have a mindset of voting only for the main parties. This mindset causes the third party to be even more unlikely to win, which feeds the cycle. The third party options are simply not practical in the current system. The voting system should be changed so that a third party is a more viable option for voters.
The Pew Research Center conducted a study about the trends in political polarization in the United States in the last two decades. They found that the number of Americans who are consistently liberal or conservative in their opinions has doubled from 10% in 1994 to 21% in 2014. They found a dramatic increase in the percentage of people in the Republican and Democrat parties who have a very unfavorable opinion about the opposing party, from 16% for Republicans and 17% for Democrats in 1994 to 38% and 43%, respectively, in 2014. More importantly, the political center has diminished from 49% taking “roughly equal number of liberal and conservative positions” in 1994 down to 39% of people being in the center in 2004 (7 Things to Know about Polarization in America).
Another article called “The Opposite of Republican: Polarization and Political Categorization” by Evan Heit and Stephen P. Nicholson talks about people’s views of the Republican and Democrat political parties. It discusses studies that showed that people view these parties as mirror images of each other more strongly than other opposing categories, such as healthy and junk food, or male and female jobs. Participants were given a list of 15 political figures or media personalities described as Democrat or Republican and were asked to rate the correlation from 1 to 7. This same process was repeated for different foods and the categories “Healthy” and “Junk Food”, and then for different jobs and the categories “Male” and “Female”. They found that there was a stronger inverse correlation between a person being Republican versus Democrat than there was for foods being healthy or junk food and for jobs being male or female.
Another explanation for why people automatically categorize others is quite fascinating. There is a cognitive framework in our brains called the “alliance detection system,” which, among other things, helps our minds make judgments of whether someone is one of “us” or one of “them”. A study by David Pietraszewski, Oliver Scott Curry, Michael Bang Peterson, Leda Cosmides, and John Tooby described and tested this phenomenon in the context of politics. They gave participants a list of political and public figures and told them to categorize the people as being more conservative or liberal, and as being more Republican or Democrat. They found that the alliance detection system is activated even when there is no “explicit cooperation and antagonism”, but “cues of support and agreement are sufficient.” (Constituents of Political Cognition: Race, Party Politics, and the Alliance Detection System) This means that people could be categorized before they make their viewpoints known, based solely on cues that party members think of as being indicative of support or opposition. Another result of the study was the discovery that people view political stance as not only a set of opinions and beliefs, but also as “predictive of the quality and nature of the relationships people will have with one another.” The us-them mentality causes people to view voting as a two-sided choice, rather than a choice between more than two parties.
Relating to the categorization with limited information is a study by Ariel Malka and Lelkes Yphtach, which explores the idea that party members are more likely to adopt a particular belief when they are told that it is associated with the person’s party. This especially applies to newly politicized issues where the political parties take a stance on an issue at the same time it is gaining the public spotlight. They told some participants that liberals were opposed to farm subsidies, some that conservatives were opposed, and gave some participants no cues. They found that when people were told that liberals opposed farm subsidies, liberals were more likely to oppose it and conservatives were more likely to support it. The converse was true for the people who were told that conservatives opposed farm subsidies, and there was no correlation between political affiliation and supporting farm subsidies when no cues were given (More Than Ideology: Conservative–Liberal Identity and Receptivity to Political Cues).
As a result of this polarization, a third party candidate is very unlikely to win. This makes the candidates that voters can vote for and have a realistic chance of winning worse. It also encourages a mindset of voting against the opposing candidate and creates a system where a vote for a third party is considered a waste. There are two ways to improve the system and present voters with better options: make available choices more practical, or make more practical choices available!
To make the available choices more practical, we need to encourage more people to vote in the primaries, so that the candidates represent the interests of more people, rather than being skewed toward the more dedicated party members. There has been a decline in the past few decades of young voters. A paper titled “Voting Trends by Age Group” has identified a decline especially in voters age 18 to 24, partly due to an increase of non-citizens in that age group. In 1964, voting rates for this age group were 50.9%, and in 2012 they had dropped to 38.0%. The voting rate by age in 2012 was 45% for 18- to 29-year-olds, 59.9% for 30- to 44-year-olds, 67.9% for 45- to 64-year-olds, and 72.0% for 65-year-olds and older. The paper also stated that there has been an increase in voting rates among older individuals.
One obvious way of getting more people to vote is by setting up online voting. This has many risks, however. Even if the communication is secure, the computer sending or receiving votes could be compromised. An article titled “If I Can Shop and Bank Online, Why Can’t I Vote Online?” explains that the level of security required for online voting is much higher than the level available for E-Commerce. More than one in a thousand online transactions are fraudulent, but presidential elections are often won by a few votes per thousand! Although making voting easier through online voting is not a viable option, there are other ways to increase voting rates.
One possible way to increase voting rates would be to run more advertisements relating to the presidential election on television. However, a study by Jonathan Krasno and Donald Green called “Do Televised Presidential Ads Increase Voter Turnout? Evidence from a Natural Experiment” found that advertisements on television have a negligible effect on voter turnout. The study references two other studies that found the opposite, that television advertisements for candidates do in fact increase voter turnout, in the order of about 5% to 10%. The paper says that these claims are debatable, since the advertisements do not directly remind people to vote, and that other techniques, such as phone calls and canvassing, are also used. These other techniques are more likely the cause for people voting when they otherwise would not have. They reference other studies that found that face-to-face contacts work best, then telephone calls are somewhat less effective, but both are more effective than direct mail. The main part of the study was the analysis of data about the number of advertisements aired in the television networks in each specific area of the country in different years, the number of eligible voting-age citizens, and the percent turnout in different years. They showed that while some states had a positive correlation between number of advertisements and percent turnout, others had no correlation or negative correlation, and that there was very little correlation overall. They estimated the cost per vote to be about $200 based on the amount spent on the advertisements and the number of votes gained, which was twice as large as the generous amount for the cost per vote of direct mail or phone calls. They also state that they did not factor in other costs such as the cost of creating the advertisement.
To make more practical choices available, we need to make more than two parties be a relevant and viable option. Voters could rank the candidates in order of preference, selecting their top choice, second choice, and third choice. Then the top choice votes would be counted, the candidate with the least votes would be eliminated, and the people who chose that candidate as their top choice would have their vote automatically move to their second pick, which would change the balance of the remaining candidates. This would continue until two candidates remain, and the candidate with the most votes would win. To illustrate, let us suppose that Alice, Bob, and Carl are running for president. For the top pick, 35% of voters select Alice, 25% select Bob, 40% select Carl, and most of Bob’s supporters would have voted for Alice had Bob not been running. In the current system, Carl would win the vote. In the proposed system, voters would have an opportunity to write down a second pick, and possibly third, fourth, and so on. Bob would be eliminated, and the people who chose Bob as their first choice would have their vote applied to their second choice. Let’s say that 80% of Bob’s supporters put Alice as their second pick and 20% selected Carl. This means that Alice would now have 55% of the votes, while Carl would have only 45%. Alice would win, which would be more accurate to the people’s preferences.
There has been a debate about whether or not we should get rid of the electoral college. The electoral college does not need to be removed for the proposed ranking system for voting to work, but the presence of the electoral college would change how it works. If the electoral college still existed, each state would count votes separately in the same way as described previously. The voters would still put down a first choice, second choice, third choice, and so on, but the votes would only be totaled up within the state. The losing candidates would be eliminated starting with the candidate with the fewest votes, and the people who voted for that candidate would have their vote reassigned to whoever they put down as their second choice. The process would continue until the top two candidates, and the candidate with the most votes would receive all of the electoral votes. If the electoral college was eliminated, the process would be much the same, except the votes would be counted through the entire United States. The process would make much more sense without the electoral college, as the same person would be dropped from the running nationwide during each round, rather than possibly having a different winner in different states. With the electoral college, when three different candidates win at least one state, the electoral votes that went to the losing candidate of the three would not be able to be easily reassigned to the second choice.
One possible weakness with this strategy would be the presence of write-in candidates. The system as I have described it so far would treat each write-in candidate separately. Since the Population of the United States is about 325 million people, and each person could submit up to three choices, this would result in a theoretical maximum of about 1 billion different write-in candidates if 100% of people vote, each person puts in three write-in candidates, and no two people write in the same person. However, only ten states and D.C. allow anyone to be written in, and their total population is only 41 million people. Most other states have some requirements to become a write-in candidate, such as needing to have a certain number of voters sign a petition. Eight states do not allow any write-in candidates (Thinking about writing in a candidate on Election Day? Read this before you do). Because of this, there would be a theoretical maximum of three write-in candidates per person in the ten states and Washington D.C. that allow any write-in, which would be 123 million write-in candidates. A computer program could easily be written to evaluate the proposed system with 325 million votes and 123 million candidates, and it would run in an order of seconds on one of the many supercomputers the government has access to. The computer described in an article by National Geographic in 2012 titled “U.S. Lab’s “Titan” Named World’s Fastest Supercomputer” says the computer operates at 17.59 petaflops, or 17,590 trillion operations per second. To do a preliminary elimination of write-in candidates that are mathematically unable to win would take a fraction of a second since only 325 million operations (multiplied by a small constant) would be needed. Then the evaluation of the remaining candidates would take less than a few seconds depending on how many candidates were eliminated. On a more modest computer, the evaluation would take a few minutes to complete. However, the real limitation is in reading the write-ins and entering them into the system. A solution to this is for the eight states and Washington D.C. to only allow write-in candidates when a certain number of people have signed a petition for that candidate.
One possible flaw with this system is that it might be too complicated for voters to understand. They would be presented with places for the first, second, and third choice, and also the different candidates. The solution to this is to allow voters to select only one candidate if they want to, and leave the second and third choices blank. The way the voting system works could easily be explained with an animation of bar graphs representing the number of votes for each candidate. The bar belonging to the candidate with the least votes would be cut into pieces of different colors representing the second choices of the voters who chose that candidate as their first choice. Then the pieces would move to the corresponding bars for those candidates who are still in the running. This would show in a simple way how the system applies the first, second, and third choices.
To test and improve the system, student government groups could adopt the process, try it out, make improvements, and measure its success. Some options to test would be how many spaces for candidates to provide on the ballot, how the system works for candidates that have similar stances on issues, and how the system works with and without primaries. The success could be measured by taking a poll of approval ratings with and without the system in place over a period of time.
The United States political party system has increased in polarization over the last few decades. This polarization causes an us-them mentality which discourages voting for a third party. The current system worsens this problem by not encouraging more than two options. There has been a decline in the number of voters in the last few decades, and if more people voted, especially in the primaries, the two main candidates for president would be better. Online voting to encourage participation would not be plausible with the current technology due to the rate of fraudulent network activity being higher than the tolerance needed when the votes are almost even. Getting more people to vote through television advertisements would be a very expensive and inefficient endeavor. Since getting more people to vote is a very difficult problem, we should try to provide more choices to voters. The voting system should be changed so that a third party is a more viable option for voters. A system where voters select their top three candidates would encourage this, as no one would be wasting their vote since it would be reassigned if their top choice did not win. The electoral college does not need to be removed in order for this system to work, but the system would run better without it. The vast number of write-in candidates could be managed by requiring a petition signed by a certain number of voters for a write-in candidate to be in the running. The system builds upon the current process so that voters can still select only one candidate if they want to, and that candidate will receive their vote. The voting system could be tested by a student government organization to make changes and measure its effectiveness. Regardless of whether or not you agree with the new system, it is good to think about the way party politics work and examine with a critical eye how well the current system represents the views and interests of the people.
Work Cited
Lavelle, Marianne. “U.S. Lab’s “Titan” Named World’s Fastest Supercomputer.” National Geographic, National Geographic Society, 30 Oct. 2012, news.nationalgeographic.com/news/energy/2012/10/121029-titan-fastest-supercomputer/.
Lu, Denise, and Kevin Uhrmacher. “Thinking about writing in a candidate on Election Day? Read this before you do.” The Washington Post, WP Company, 3 Nov. 2016, www.washingtonpost.com/graphics/politics/2016-election/write-in-votes/.
Krasno, Jonathan S. and Donald P. Green. “Do Televised Presidential Ads Increase Voter Turnout? Evidence from a Natural Experiment.” Journal of Politics, vol. 70, no. 1, Jan. 2008, pp. 245-261. EBSCOhost, search.ebscohost.com/login.aspx?direct=true&db=a9h&AN=38604362&site=ehost-live.
“Voting Trends by Age Group.” Congressional Digest, vol. 96, no. 1, Jan. 2017, pp. 12-14. EBSCOhost, search.ebscohost.com/login.aspx?direct=true&db=a9h&AN=120541607&site=ehost-live.
Jefferson, David. “If I Can Shop and Bank Online, Why Can’t I Vote Online?” Verified Voting, 22 Sept. 2012, www.verifiedvoting.org/resources/internet-voting/vote-online/.
Doherty, Carroll. “7 Things to Know about Polarization in America.” Pew Research Center, Pew Research Center, 12 June 2014, www.pewresearch.org/fact-tank/2014/06/12/7-things-to-know-about-polarization-in-america/.
Bless, Herbert and Norbert Schwarz. “Context Effects in Political Judgement: Assimilation and Contrast as a Function of Categorization Processes.” European Journal of Social Psychology, vol. 28, no. 2, Mar. 1998, pp. 159-172. EBSCOhost, search.ebscohost.com/login.aspx?direct=true&db=a9h&AN=11830071&site=ehost-live.
Malka, Ariel and Yphtach Lelkes. “More Than Ideology: Conservative–Liberal Identity and Receptivity to Political Cues.” Social Justice Research, vol. 23, no. 2/3, Sept. 2010, pp. 156-188. EBSCOhost, doi:10.1007/s11211-010-0114-3.
Pietraszewski, David, et al. “Constituents of Political Cognition: Race, Party Politics, and the Alliance Detection System.” Cognition, vol. 140, July 2015, pp. 24-39. EBSCOhost, doi:10.1016/j.cognition.2015.03.007.
Sønderskov, Kim Mannemar and Jens Peter Frølund Thomsen. “Contextualizing Intergroup Contact: Do Political Party Cues Enhance Contact Effects?.” Social Psychology Quarterly, vol. 78, no. 1, Mar. 2015, pp. 49-76. EBSCOhost, doi:10.1177/0190272514560761.
Heit, Evan and Stephen P. Nicholson. “The Opposite of Republican: Polarization and Political Categorization.” Cognitive Science, vol. 34, no. 8, Nov. 2010, pp. 1503-1516. EBSCOhost, doi:10.1111/j.1551-6709.2010.01138.x.
Heit, Evan and Stephen P. Nicholson. “Missing the Party: Political Categorization and Reasoning in the Absence of Party Label Cues.” Topics in Cognitive Science, vol. 8, no. 3, July 2016, pp. 697-714. EBSCOhost, doi:10.1111/tops.1220