Numerous pathologies: the fascinating dysfunction of the US electoral system.
Last spring, before the coronavirus and the general presidential election began to consume our lives, the Democratic party held a series of primaries to determine who would be their nominee for the general election. There was no real Republican primary this year, so I will focus on the Democratic primary. These primaries – and also the general election that they were instrumental in – are an interesting example of the different pathologies that can plague voting systems. Notably, the primary (and thus the general) are distorted by both the pathology of strategic agenda-setting, and the pathology of strategic misrepresentation.
The Democratic primary begins with a primary in Iowa, then a couple other contests in smaller, not particularly representative states – New Hampshire, South Carolina, and Nevada. Then, a number of contests are held on “Super Tuesday,” in March. At this point, even though not all of the states have held a primary, the nominee is basically already decided. By the time that New York votes in June, New York voters could basically be called disenfranchised; their votes will most likely not change the direction of the primary at all.
The phenomenon of strategic agenda-setting is when changing the order of the singular elections would change the final outcome of the election, in this case the primary. This is a pathology that affects elections based on majority rule in which there are multiple contests to determine a final ranking. New York and California are populous states with a number of voters who prefer more left-wing candidates over moderates. If those two states were to go first, the primary would be shaped by those more liberal voters, instead of the more moderate, mostly white voters of Iowa. The first states to vote matter not only because of “momentum,” where a candidate winning makes people think they can continue to win and thus vote for that candidate, but also because as the primary continues, the number of candidates decreases. Running a political campaign is expensive, and if a candidate does not think that there is a path for them to win, they will drop out of the primary and save their money (and dignity).
A feature of the 2020 Democratic primary, as well as the 2016 Republican primary, was the vast number of candidates. With such a large number of candidates, a number of them will necessarily occupy the same general niche in terms of policies. For example, Kamala Harris, Amy Klobuchar, Pete Buttigieg, and Joe Biden were all moderate candidates, and thus vying for the voters who prefer moderate candidates. Elizabeth Warren and Bernie Sanders were likewise fighting for the voters who prefer more left-wing candidates. Since all four of the moderate candidates are technically appealing to the same kind of voter (although each of them appealed in some way to distinct demographics) having so many candidates actually disadvantages the frontrunner of the moderate candidates; the moderate vote could end up split just enough to weaken the frontrunner and let a left-wing candidate gain a larger plurality. Biden won decisively in South Carolina – where the majority of the Democratic voters are Black, a demographic that was favorable toward Biden – and his moderate opponents dropped out to avoid such an event. If the states had been ordered differently, causing his moderate opponents to have more hope and stay in the contest, that could have continued to split the moderate vote and made the primary ultimately have a different conclusion.
Additionally, the primary is only the prelude to a general election, in which voters of all parties can participate. This makes the primary not only a simple ranking of the majority’s preference, but a matter of strategy, where voters are also trying to pick a candidate who can win in the general election. This can lead to another pathology: strategic misrepresentation, where voters lie about who they actually prefer in order to get a desired outcome. This could be purposefully voting for a spoiler to weaken a candidate who looks strong – or voting for a candidate who is not actually their top choice in the hope that that candidate will be more likely to win in the general election, aka is more “electable”. That is likely what happened with Black voters in South Carolina: the New York Times quotes a Black voter who says, “…we’ll back Biden, because we know who white America will vote for in the general election in a way they may not tell a pollster or the media.” Clearly, some of the people who voted for Biden were not doing so because they liked him or thought he was the best choice to be president; they were strategically misrepresenting their preferences in order to elect someone that they thought could win in the general.
This primary process may or may not end up choosing the candidate that the majority of Democratic voters actually prefer. It is heavily biased toward the early states, which each have their own demographic quirks. Votes in states that vote much later mostly do not matter. The sequence of elections may lead to candidates dropping out who could have changed the race if they had stayed in longer (although it is difficult if not impossible to prove such a counterfactual). Additionally, voters have an incentive to misrepresent their true preferences in order to achieve their desired electoral outcomes. We have ended up with a system heir to not only the pathology of elections based on majority rule, but also the pathology of elections based on positional voting. No democratic voting system is without flaw – Arrow’s Impossibility Theorem says that in an election with multiple alternatives, the only system that avoids these pathologies is a dictatorship – but surely it is possible to have a primary process that is less flawed.
A defense of our current system is that it makes sense for specific states to have more say in the choice of nominee, because the electoral college, the system that elects the president, ultimately rests on the results of a few states that either party could win, and not on the national popular vote. The electoral college is a system that is somehow even more dysfunctional than the primary process – at least the primary process ensures that the eventual nominee gets the most votes! The electoral college is a “democratic” system that can award the presidency to the candidate who lost the general election. It is a voting system that cannot even reliably serve its purpose: to rank the candidates according to the preferences of the people of the United States. The dysfunction of the primary system and the general election system reinforce each other; the primary system is biased toward a few states because it is necessary for the eventual nominee to win a few key states and thus the electoral college, and the general election magnifies that dysfunction, making it so that the elected president may not even win a majority of the votes.
The deep flaws of our systems are interesting from a detached perspective, but terrifying for a person who wants to live in a truly democratic country. I can only hope that the way we elect our presidents will be amended to be less pathological within my lifetime.
sources:
https://www.usnews.com/news/blogs/data-mine/articles/2016-01-21/why-the-iowa-caucuses-matter-to-the-2016-presidential-election
https://www.theguardian.com/us-news/2020/mar/04/joe-biden-jim-clyburn-endorsement-super-tuesday
www.nytimes.com/2020/02/29/us/politics/black-voters-south-carolina-primary.html
https://www.vox.com/2020/2/28/21153287/delegates-democratic-primary-2020-math
https://www.vox.com/2020/1/27/20686864/democratic-primary-calendar-2020-iowa-super-tuesday