The writer, who appears to be a political scientist, doesn't know much about political campaigns, or at least how political campaign polling works. This wasn't

The writer, who appears to be a political scientist, doesn't know much about political campaigns, or at least how political campaign polling works. This wasn't a push poll. A push poll wouldn't waste time asking the voter for demographic information or about likelihood to vote - they would get to the negative info they want to spread within a few seconds of the call connecting. Instead, this sounds to me like the common format for a campaign poll that is used to measure the effectiveness of negative messages (so, calling a small # of voters for the purpose of fine-tuning negative ads, etc). If he had kept listening, he likely would have heard the pollster test negative messages on Clinton as well (this is also common - you're trying to figure out what messages work against you at the same time that you test your opponent's negatives). I imagine the Sanders campaign is conducting similar polls (almost any campaign with more than a couple of million dollars raised would conduct a poll like this) and, in fact, this very poll could have been coming from the Sanders campaign.