After its disastrous polling attempt in the 2012 elections, Gallup has decided to get out of the election predicting business. Harkening back to some of the off base predictions: Dick Morris, “This is going to be a landslide.” Karl Rove, “At least 279 electoral votes.” While Gallup was off the mark with the widest margin, the vast majority of pollsters picked Romney to win leaving Rove and Morris with egg on their face and the GOP establishment searching their souls for what went wrong. In hindsight, perhaps promoting neocons for the past two cycles was not the best choice for a country that was weary of war and foreign interventions.
In the 2014 elections there were a few predictions that the GOP might pick up 53 seats but that was with the rosiest of presuppositions. Most polls considered the most competitive seats to be toss ups. No one believed that this year would be a Republican avalanche. The polling was biased towards democrats but the result was that the GOP garnered 54 total Senate Seats and picked up 13 new congressional seats in a sound shellacking of the party of such liberal ideals as open borders, gun control and abortion rights. Even traditionally democrat states flipped in North Carolina and Louisiana. The conservative surge propelled David Brat to defeat the establishment Republican Majority leader, Eric Cantor in the primary on a shoe string budget. This primary challenger crushing should have been a warning sign that the conservative tide was turning. The surveys underestimated how the GOP would do in close races and polls have not made corrections for this years elections either.
Then we had the more recent Kentucky governors race with GOP candidate and Tea party favorite Matt Bevin who polls said was trailing Democrat Jack Conway by 7 points. Never did the polling have Bevin ahead yet he won by a huge 9 point margin. Another fail for the pollsters.
According to an article in Reason Magazine by Stephanie Slade, “pollsters try to figure out who will show up on Election day by asking respondents how likely they are to vote. . . People it turns out , are highly unreliable at reporting their future voting behavior. “ It’s getting harder to find the right combination of polling methods, direct calling, online polling, social media trends, that will predict accurate outcomes. Fewer respondents are participating and another problem is that polling companies are trying to save money and follow the herd of other pollsters who follow more rigorous methods. They “fudge” their numbers to trend with Gallup and other big name polling institutions. Trends tend to be foolish and those who follow trends often look like fools.
In Iowa for example the polls typically call older voters(50 and up) and many millenials have cell phones which are left out of the equation. Rand Paul noted, “We think that a lot of young people are not included in polls, college students and younger people with cell phones, and we think that’s where our great strength is.” Rand supporters tend to be the most die-hard of all the candidates and it shows in their online polling dominance. On Facebook Rand Paul has over 2 million likes, third only to Ben Carson and Trump( many of whose likes came from suspicious locations in Asia). Paul is also third in Twitter followers and Instagram according to an article on Breitbart.com.
Another problem of polling is bias. If the pollsters can indicate lack of support for a candidate then donations and support typically dry up as well. Thus biased pollsters can target a candidate for marginalization by underplaying their support. With over 1,000 precinct captains in Iowa alone and numerous endorsements and college clubs, Rand will over achieve once again. As you may suppose, the establishment is not going to help a candidate who wants to roll back our aggressive foreign policy, audit the Federal Reserve and cut back on their crony spending plans, but if the people have their say, they will vote for the candidate who puts Americans first, not the corrupt Washington Machine.