Overall this cycle the polling has been pretty good across the early voting states. This isn’t to say all of the polling was credible and certainly some surveys, such as those showing Joe Biden down or tied in South Carolina a day or two before the Palmetto State’s primary were just wrong.
Looking at the chart below we see that most candidates performed about 3-to-4 points from where polling had predicted, which given the fact that the margin of error in most of these surveys was in the 3-to-4-point range means that in general, the polling was solid.
Now, let’s talk about Super Tuesday. For the most part, the polling that took place immediately before Super Tuesday indicated that overall, we should have expected two candidates to emerge from the pack: Sanders and Biden. Their dominance was clearly predicted in polling. Not only were they the only two candidates predicted to win states, but they also were never predicted to perform below the 15 percent viability threshold. This turned out to be correct.
But much of the polling conducted prior to Super Tuesday failed to accurately capture the degree of success Biden experienced this week. In every state where Biden was predicted to win (AL, AR, NC, OK, and TN) he won by a much bigger margin. More shockingly, in several states where Sanders was predicted to win, Biden came out victorious (ME, MA, MN, and TX). But we cannot fault polls that were conducted before South Carolina and before several candidates dropped out — and subsequently endorsed Biden — for not capturing Biden’s last-minute surge.
Now, most pollsters are calling voters over a minimum of a three-day period, which reduces bias. So their ability to gauge voters’ opinions in Super Tuesday states following the influential South Carolina vote was basically zero. The strength of Biden’s win in South Carolina, coupled with the short times period, and multiple candidates dropping out, makes the underestimating of Biden’s support understandable.
Since the polling was largely conducted before South Carolinians voted, they were unable to capture just how influential Biden’s win was in creating momentum into the Super Tuesday Primaries.
The result led him to increase his support by 4 points in CA, 36 percent in ME, 126 percent in MA, 21 percent in TX, and 12 percent in VA. His win had compounding effects that were difficult to capture within such a short period of time between primaries. The South Carolina win alone was great to reignite Biden’s campaign and revitalize voters who had doubts regarding his electability and many polls reflected this.
But it also led fellow moderate Democrats Pete Buttigieg and Amy Klobuchar to drop out of the race and endorse Biden instead. The polls for Super Tuesday were conducted in a different environment than actually occurred—one where there was not a clear moderate leader in a wider field—so it’s not surprising that they didn’t completely reflect the swell of support Biden would receive.
Given Biden’s electoral surge, the media attention he’s receiving, and the narrowed field, we hope to see new polls in the states that are coming up — Idaho, Michigan, Mississippi, Missouri, North Dakota, and Washington on March 10th — in order to fully grasp what impact the results of Super Tuesday will have for the next phase of the cycle.
Not sure it will happen especially in smaller population states like Idaho and North Dakota, but here’s hoping.
Stefan Hankin is the founder of Lincoln Park Strategies.