The Problem with polls
With only two weeks separating today from this year’s midterm elections, the prospect of widespread predictive polling is inexorable. For decades, polling in the American political sphere has served as an overwhelmingly reliable way of predicting elections.
In years past, pollsters could expect anywhere from 80-90 percent of poll targets to respond and share their political views.
Today, though, in the midst of what some consider to be one of the most historically polarizing periods in America, the reliability of polling sits in question.
In 2017, The New York Times’ Upshot model, which determines the correlation between predictive polls and election outcomes, determined that the likelihood of a polling misfire today is about as likely as a baseball strikeout. That’s close to a 20 percent inaccuracy rate.
This year in particular, though, we’ve seen an extensive decrease in the accuracy of political polling across the board, particularly facing the Democratic Party as Trump’s unpopularity spearheads newfound anger and excitement within the party and its voters.
For example, let’s take a look at this year’s Florida gubernatorial primary.
For the entirety of the pre-election season, Democratic primary frontrunner Gwen Graham stood safely ahead in the large majority of predictive polls. Tallahassee Mayor Andrew Gillum, however, stood reliably in third or even fourth place in most polls for the same race.
The polls, as you presumedly know, were wrong.
To the surprise of pollsters and citizens across the state of Florida, Gillum acquired the nomination by almost 45,000 votes. So why were the polls wrong?
As explained by Evan Horowitz of the Boston Globe, “Hidden behind those definitive-looking, topline poll numbers is a set of hard-to-test assumptions about who is actually going to show up to vote and who — among all the registered voters in a district — actually counts as a likely voter, the foundation of most polls.”
Essentially, while orderly polling practices may consider a voter’s voting history, the prospect of anger and enthusiasm at the polls will never fully be accounted for.
In this election, pollsters significantly underestimated the magnitude of youth and African-American turnout at the polls. Historically, African-American voters and youth voters tend to stay home during midterm-related elections, but as a result of the political excitement within the Democratic Party’s base, Gillum, who amassed a large amount of support from youth and African-American voters, contradicted history.
Should I bother looking at polls?
Yes! Although the legitimacy of polling has been somewhat questionable in years past, multiple political organizations and well-known publications have altered their polling methods to make them more efficient.
For example, Nate Silver’s Fivethirtyeight prediction service, acclaimed for its impressive accuracy in predicting house, senate and presidential races, utilizes public polling, fundraising data and historical trends to predict the outcome of elections.
Here are some of the most reliable and innovative polling websites and organizations that you should check out:
Photo provided by leedsjld.com