Electoral Models: Why Past Successes Shouldn’t Dictate the Future
The linear tyranny of time is a major problem with which political scientists must grapple.
Public opinion both influences and is influenced by the metrics scientists use to capture attitudes towards hot button topics and assess the competitiveness of local, state and federal races.
But, all the tools used to measure current races are based on successful algorithms from past cycles. And while there are many fitting adjectives for American politics, many of them pejorative, “static” is not one.
Political science has not adjusted well to the digital age. Perhaps the most obvious example of this in 2016 is the much-contested method Fox News, which will host the first party debate in August, is using to decide which of the plethora of candidates will participate- the top ten as ranked by opinion polls.
Yet, as any honest political scientist will tell you, this is a seriously flawed science, almost impossible to rid of bias. As a result, a few of the candidates, have announced their intention to focus on their ground game in states with friendly demographics instead of pandering to national media conglomerates.
Already discovered by political activists, social media has risen as another alternative for politicians shunned by the press. And though the forest can sometimes get lost for the trees, underdogs like Jim Webb, who advocates a lot of his policy vision through his personal Twitter, and more savvy candidates like Donald Trump, whose followers are treated to some very interesting if factually challenged quips directed at his competitors, can be heard, even if sometimes it is only to an echo chamber of ideological devotees.
Unfortunately, these are the outliers in candidate interaction. Traditional methods of prediction still dominate media, and thus heavily influence elections.
But, how accurate are some of these in a world where democracy is increasingly digitized and personalized?
Statistical wunderkind Nate Silver’s website, FiveThirtyEight, is a veritable treasure trove for numbers geeks. He is one of the most accurate predictors in politics.
FiveThirtyEight currently features an interactive page detailing the history of endorsements as a predictor of presidential politics, historically very accurate. Measured by distance from the first-in-the-nation Iowa caucus forward, the model shows the contender with the most number of endorsements from governors, Senators and congressman having a greater chance of winning.
By the numbers, this looks likely to hold true for assumed-Democratic nominee Hillary Clinton, who has already racked up a stunning number of endorsements from her fellow party leaders and who holds a consistent 30-point lead over her running mates in polling.
But what about the much more competitive Republican field? Though endorsements are a lot more scarce on the other side of the ideological aisle, Chris Christie leads, followed by Jeb Bush and Rand Paul.
A recent national poll conducted by Fox News, however, shows Donald Trump as the leading Republican contender, followed by Scott Walker, then Jeb Bush, then Rand Paul.
This discrepancy could be a result of many factors, but is most likely a product of the number of contenders. Bush, like Clinton, is the establishment favorite. However, conservatives, angry at having their concerns cast aside by those same establishment figures, are angry, and now have more ways to vocalize this.
The point: statistical models are constrained to past patterns, and are therefore fallible.
American media has a tendency to rely on traditional mathematical models as a framework for election coverage, often couching their projections in an inevitability hat has the power to move public opinion.
Yet, new technologies are broadening the ability for average Americans to participate and shape the political landscape, as can be seen in the state of the Republican primary.
Minority voices can have impact, meaning Americans who see their views cast as fringe by bigger societal actors should not be so quick to self-disempower themselves. The models only become inevitable when a plurality convince themselves this must be the case and, rather than vote their conscience enslave themselves to a “lesser of two evils” rationale.