Over the course of a poll’s life – its production, analysis, and consumption – there are circumstances that can reduce its value.
Not only are there are so many variables that contribute to an “efficient” poll, but the reporting of polls (specifically, candidate election polls) sensationalize elections and do not offer voters meaningful information. There is no clear virtue in polling other than for strategic purposes for campaigns or pure curiosity.
The timeline of the poll is as follows: In every election year, NBC, Reuters, Bloomberg, LA Times, and others conduct polling to determine which candidates are ahead with various demographics on specific issues. The media then looks at the results of these polls and chooses which ones are newsworthy — and reports on only those polls. And as consumers of the news, we use these polls to check in on how the election is, in theory, going.
Polling itself is a risky business. The Pew Research Center recently published an article describing how it is all done: pollsters conduct their polls, asking a series of question to gauge whether respondents are “likely voters.”
More likely than not, all the people who respond to the polls will not actually end up voting, and those who do can easily change their minds between the time of the poll and Election Day.
There is a margin of error in every poll that indicates that the results can waver by a few percentage points depending on the sample size.
The bigger the sample size, the smaller the margin of error, and vice versa. Because a pollster cannot poll the whole nation, they have to settle for a smaller sample to provide an idea of what people are thinking. And because it is only an idea, polls leave room for error or misinterpretation.
Despite all the obstacles to make and execute a successful poll, most polls have been historically accurate. Shout-out to all the pollsters out there working hard. But does their hard work and accuracy matter if the media reports on them in a skewed manner?
The media is known for covering elections like horse races. The Washington Post conducted a study of the 2008 election comparing the polls’ results with how and what the media was reporting. The study showed that the media spent more time on reporting polls that suggested that the race was tighter than it actually was.
Since August 2015, RealClearPolitics, a news and polls aggregator, has reported Hillary Clinton ahead of Donald Trump essentially the entire time for the general election presidential race. FiveThirtyEight also rates the chance of Clinton winning as very high.
But the media is still reporting on this election like it is an incredibly tight race. “Think Hillary Clinton Will Win in a Landslide? Don’t Bet on it,” and “Why Clinton can’t shake her private email scandal” are just a two recent examples of the coverage. If the the Washington Post conducted a similar study on this election, the results would be very similar.
It is possible that polls could have a bandwagon effect on voters–they could see a poll and feel compelled to vote with the majority.
According to an article published by Slate, having a lead in the polls can only give a candidate a four-to-five point bump among undecided voters. There really is not much incentive in being ahead in the polls because it does not affect what the voters do and it also is not accurately reflected in what the media is reporting.
Polls can be helpful and interesting when reporting data on demographics and issues; but when they are exclusively reporting that whichever candidate is ahead by however many points, it just contributes to the sensationalism of election season. Polls do not offer any information to help voters make informed decisions, and, if anything, how the media reports on polling just further obscures the actual issues at play in an election.