Paul Sherman: Pointless polling



By Paul Sherman  On  April 11th, 2013

What’s up with the recent polling for the 2016 Presidential election? It seems that there's a poll coming out nearly every day about the 2016 race. From potential candidates to projections, polling organizations and media outlets have decided to devote some time to this story. Representatives of Gov. Rick Perry (R-Texas) have even announced that they are going to decide by this June about his presidential bid. In all honesty, it makes me sick to my stomach.

Ever since the end of the 2012 election cycle, media outlets such as MSNBC and CNN have been talking about who will run in 2016. To me, it’s ridiculous that we're beginning to have these discussions now. The election is more than 1,306 days away! This type of polling isn't constructive, as there are more important problems that we should be addressing.

While there have been some cases in which the polls accurately picked the candidates, most of them have not done so. According to a Gallup poll from July 2009, “72 percent of Republicans and Republican-leaning independents have a favorable opinion of Palin, compared with 56 percent for Romney and 59 percent for Huckabee.” Obviously, neither Palin nor Huckabee ended up making it past the primaries. For the 2004 election, polls projected the top three Democratic candidates to be Hillary Clinton, former Vice President Al Gore and Sen. Joe Lieberman (I-Conn.), yet John Kerry won the Democratic nomination. These polls may tell us how voters feel about the two political parties now. However, a lot can happen in four years to make voters change their minds.

Part of the problem is statistical bias. Polls won't be accurate if they don't find a representative sample of the population. This is difficult since it would be expensive to find a large enough sample that can be representative of the entire American population. However, if this condition isn't met, the amount of bias increases and the accuracy decreases. Polls can also suffer from what's called response bias. This occurs when pollsters attempt to predict what would have been the response from participants who don't respond to the questions. As an article in The Economist describes, when pollsters have to call people who don't have landlines or don’t pick up their phones, for example, “they are left guessing about how to weight their views," which can create bias and “lead to lots of mistaken predictions.”

However, there's a place for these early polls. For lesser-known candidates, such as Cory Booker, mayor of Newark, New Jersey, it helps them gauge their support and figure out the actions they need to take to become more prominent in the national dialogue. Not only that, it can help us to understand American opinions of the two political parties, Congress, the presidency and so on. In an article on Slate, John Dickerson said that talking about presidential races can “excite ambition, which causes politicians to take risks, which can start conversations about important ideas. Our day-to-day political life is small, but in the presidential conversation, it's still possible to imagine bigger things.”

There are also other problems in polling, such as the “likely voter model” and the push poll. As Frank Newport, editor-in-chief of Gallup, explained in an article by Ezra Klein in the Washington Post, the likely voter model “takes into account changes in the response to questions about how closely they’re following and how enthusiastic they are,” he said. “It’s not just capturing underlying movement — it’s representing changes in enthusiasm.” In addition, the push poll has been used, which is when a small sample of about 300 to 1000 people are used for solely measuring how information can change public opinion, according to Stu Rothenberg of Roll Call. Because of these problems, it's hard to tell whether or not polls accurately reflect the views of the American public.

So, in the next year or so, stop talking about 2016. America is facing more pertinent issues right now. It's difficult for Congress to even pass fiscal proposals that will keep the government open for business. In the grand scheme of things, polls for the 2016 election are just another distraction that America doesn't need.

Paul Sherman can be reached at pausherm@umich.edu.


Printed from www.michigandaily.com on Wed, 23 Jul 2014 11:45:01 -0400