How to Read the Polls Like a Pro

Just before we went to tape the latest episode of the San Diego Decides podcast, former Obama adviser David Axelrod had a fortuitous tweet:

Polls are so numerous at this point that results vary greatly and everyone can find one they like. Yet all are covered as if they're right!

— David Axelrod (@davidaxelrod) September 14, 2016

During election season, various polls get blasted in people's faces like they're being hurled from a T-shirt gun. Making sense of them all – especially when they sometimes seem to contradict one another – is tough.

In this episode, Ry Rivard and I run down some of the many things to consider when reading a poll: Who was paying for it, who was surveyed and how (did the interviewers use cell phones? Did they have Spanish-speakers available?) and what questions were actually asked. We also talk about the rise of online polling – a method that was considered kind of a joke as recently as a few years ago but is growing more sophisticated at a rapid clip.

And we talked with John Nienstedt of Competitive Edge, who's been polling San Diegans for decades, about lessons he's learned over the years.

Nienstedt said that even subtle changes in the wording of a polling question can produce significant differences in results.

He pointed out that in 2015, the Chargers paid for a survey that asked: "Do you favor or oppose the city and the county spending $400 million to build a new NFL football stadium in the Mission Valley area of San Diego?" About 61 percent of the respondents said they opposed that plan.

"I pointed out when this was touted to me – the question itself, if you read it carefully, what is a respondent supposed to say? Because you're putting the emphasis on the city and county spending $400 million – that's what you're responding to," Nienstedt said. "You're not responding to a package or a proposition or an initiative, you're responding to whether they should spend that money. The default is gonna be no."