Three quarters of the Bernie Sanders campaign’s doorstep canvassing was wrong
A piece about the importance of text messaging in (US) political campaigning had this eye-catching statistic buried deep within it from Cole Edwards, a former Distributed Organizing Director for Bernie 2020:
We gathered thousands of [voter] IDs via doors, texts, and through our campaign’s relational app. Then, we hired an independent pollster and polled hundreds of IDs gathered through each method.
What we found was pretty astonishing: it turned out IDs gathered through knocking doors didn’t really hold up. Only 25% of respondents who were canvassed at the door actually identified as Bernie supporters in the poll. We assumed this was because they just wanted us off their porch, but it’s hard to say for sure.
Three out of four people canvassed on the doorstep as Sanders supporters turning out not to support him? That’s a remarkably low figure and suggests that weaknesses in his grassroots campaign organisation should have got rather more attention at the time.
Sign up to get the latest news and analysis
"*" indicates required fields
Over-enthusiasm of inexperienced canvassers? Or people changing their minds and this not being picked up?
Given that we have “voter intention” within Connect plus Marked up registers would there be any benefit in following up on the accuracy of our results?