Tuesday, November 8, 2016

Big Data and the 2016 Election

Given that the next president of the United States will be decided tomorrow, I think it’s as good a time as ever to discuss the impact that big data has had on the election process. We first saw data analytics being used in a significant way during Obama’s 2008 campaign. They hired a small team of Silicon Valley recruits to stretch the fundraising efforts of the Democratic National Party and the team had remarkable success given their meager resources. In this 2016 presidential election, we have seen political consulting firms using complex models and advanced algorithms in order to determine the best way to sway and influence voters. The trick is to convince as many undecided voters as possible without alienating any of the voters you have on your side of the fence. Prior to the use of big data, public sentiment was gauged through the use of telephone polling, which is still a very widely used polling practice. This polling technique is a very refined/controlled scientific process, however, you are limited to a number of issues you can ask individuals in the sample group. Predictive analytics, on the other hand, can play out a head-spinning number of outcomes given a vast amount of variables so that you can get an idea of what candidates need to say/do in order to get the maximum swing voters to your side. This process takes the meaning of a “calculated political move” to an entirely new level.   So what does this mean for us citizens as we approach decision day? Is this technology proof of a rigged election system? For me, personally, I see it as business as usual. The use of big data analytics is focused on swaying large swathes of voters, many of whom are undecided. As long as the marketing campaigns are honest and inform voters, I believe they open up room for much needed political discourse in our daily lives. I have had several conversations with people about Trump/Clinton banner ads that have popped onto my computer and, while extremely annoying, they empower me to give a little more thought towards which candidate I want to support. What has been a significant drawback of all of this marketing/media attention, however, is that people are feeling overwhelmed by the sheer volume of over sensationalized news stories. I believe a more effective use of data analytics would be to cut through all this noise and provide individuals with information that they can talk about with the people in their lives and ultimately act upon as an informed citizen. Of course, each parties campaign is going to be biased, but I would rather get targeted by an ad that informs me on an issue that I care about then listen to the overly sensationalized rhetoric that has been playing nonstop on all of the big news stations for the past several months. 


1 comment:

  1. Since the elections have recently occurred, Ryan picked an excellent topic to discuss. Can big data inform the average American citizen to be more intellectual in deciding who to vote for? Of course, we have the Democratic and Republican point of views; but what about the moderates? To counter problem with moderates, these parties use algorithms to persuade the moderate voter. Are there any downsides to using big data in the election? Absolutely, let’s take a glance from another angle.
    The data that forecasted Hillary’s victory was obviously a fluke. A few of major forecasters, The New York Times Upshot and the Princeton Election Consortium, had predicted Secretary Clinton to win (ranging from 77-99 percent). Real stuff, huh? This leads to my next point… What does it mean for these forecaster businesses? First off, since the margin of error is tremendous, it will probably mean New York Times Upshot can be proven to be an invalid resource for the election. When the next election occurs, people may be reluctant to use these political major forecasters because of the upset brought by Trump.
    How was there a flaw with the data then? Big data in the political realm can come with a few trade-offs. For instance, it can sometimes be seen from one perspective of politics as well as fail to address other issues (missing context in the data). In the NY Times article, Google Flu Trends had to predict whether the 2012-13 flu season would be tremendous based on trigger words typed in the Google Search. Based on the Google Search, many people seemed to be worried and anticipating the flu season to be a huge threat during 2012-13. On the other hand, it turns out the data overestimated the people who was affected by the flu.
    This applies to Trump’s victory. Just because it is forecasted by New York Times Upshot or other forecasters, it doesn’t mean it will necessarily come true. Politics is an unstable subject in that anything is possible. Who knows? Maybe Trump can run and win again despite the odds against him



Note: Only a member of this blog may post a comment.