With the midterms fast approaching there is a question on every political analysts and pundits mind, will the polls be accurate this time? In electioneering, polls are still the only real way to measure voter intentions going into any democratic contest, yet in the recent past beginning with the Brexit referendum they have failed to accurately predict elections and referendums on both sides of the Atlantic.
In the 2016 election the vast majority of polls suggested Hilary Clinton would be the next President of the United States by a handy margin, yet not only did that not come to pass, Trump won the election by 77 Electoral College votes.
For example, Michigan was considered practically in the bag for Hilary Clinton during the 2016 election cycle. Polling numbers less than a month out from Election Day had Donald Trump’s campaign lagging up to a huge 11.4 points behind in the race. On Election Day however Trump emerged victorious by 10,704 votes.
Michigan Democrats were absolutely shocked that the Clinton campaign and the DNC listened to the polls predicting an easy victory, over their own on the ground experiences which painted a very different picture. These battle-hardened political operatives very much knew that Michigan was a live race in the battle for the Presidency, yet the Clinton campaign chose to stand behind their “model” based around pre-selected data.
In 2016 the Democrats put their faith in the polling, in the data and left themselves putting all their eggs in one basket. On November 8th 2016 the bottom fell out of that basket and left them not only with egg on their faces, but picking up the pieces in a world in which Donald Trump would now be President and the Republicans would control both the house and the Senate.
Fast forward to today, some of the same polls that predicted the Clinton campaign handily winning the Presidency are now projecting a strong possibility that the Democrats will retake control of the house. The Real Clear Politics Generic Congressional Vote polling currently gives the Democrats a mammoth 5.7 point lead over the Republicans. Democrats are rejoicing at the possibility of Congress going blue and finally having a means to check Donald Trump’s agenda politically.
Yet in this renewed faith in polling there is danger that the mistakes of the recent past may once again be repeated, with potentially dire consequences for the Democratic Party all over again.
Since the 2016 Presidential election, political discourse has become even more polarized, seemingly by the day. As the two sides drift further and further apart, regular every day voters feel less and less comfortable sharing their political opinions with those that don’t share their particular values.
People are increasingly keeping their own counsel when it comes to political matters, with some even going as far to outright lie or deceive others about their voting intentions in order to maintain their relationships and communities.
Trump supporters and Republicans are often vilified by prominent individuals or media commentators and this has led to them increasingly dropping out of political discourse but increasingly motivated to get out and vote come November.
In all these pieces of commentary there is often the common thread that you are not a good person, or a good American if you want to vote Republican. With all this negativity surrounding anyone voting for the GOP of course some people are going to be reluctant to share their opinion if they want their state to go or remain red.
As some Republicans and Trump supporters increasingly go underground, this potentially makes the pollsters job that much harder. If people that are interviewed can’t or won’t be truthful about their genuine political allegiance then the polls will become increasingly skewed towards pointing to victories for the Democratic Party.
If Democrats hope to win back the House and make a dent in the Republican majority in the Senate they must be careful not to repeat the mistakes of the recent past and this time listen to those with genuine sources on the ground, rather than relying solely on big data to give a complete picture.