"SHOCK POLL: 46% Think Hillary Should Suspend Her Campaign..." reads the headline on the Drudge Report. Whoa, you think. The former secretary of state's political troubles in her 2016 campaign are way worse than even you thought!
I mean, maybe. The e-mail server issue is a problem for Clinton as is the not-so-good way she and the campaign have handled it. But this headline -- and the Rasmussen Reports poll that it links to -- doesn't prove that half of the country wants to get rid of Clinton. In fact, the poll proves very little.
Let's go through the problems with the poll -- in no particular order.
1. It is of all likely voters nationally
Does it surprise you that Republicans think Clinton should suspend her campaign? It shouldn't. If you asked "Should Hillary Clinton immediately move to Mars?" my guess is that loads of Republicans would agree with that statement. I am certain Hillary Clinton and her campaign are not cowed by the fact that lots and lots of Republicans think she should suspend her campaign.
2. It is really badly worded, Part 1
So, here's the question in, um, question:
So, there is a "should" option. But no "should not" option. Years and years of polling and studies of polling tells us that if you have a "yes," you also need a "no" -- that lack of balance in question wording can push opinion far more to the "should" option since the "should not" option isn't presented at all. The better way to frame the question is: "Do you think Hillary Clinton should or should not suspend her campaign...."
3. It is really badly worded, Part 2
The way the question is asked presumes that Clinton is in some sort of legal trouble that might lead to her having to suspend her campaign. She's not. She hasn't been charged with anything at all and she's not a reported targeted of the investigation. And not only that, but there is no counter-argument posed in the question. You have one option to choose from: Are Hillary Clinton's problems with her e-mail server sufficient to cause her to leave the race? That's it.
4. The survey’s methodology is not the most rigorous
Rather than a poll using a random sample of cellular and landline phones, Rasmussen’s automated surveys only dial landline phones and supplement this with a non-random sample of adults reached online. Results are weighted to a target population of likely voters, but the firm's surveys have struggled with accuracy and missed the mark in recent elections.
When you ask bad or vague questions, the data produced from those questions tends to be, wait for it, bad and vague. What the Rasmussen poll has "proven" is that 46 percent of likely voters are ready to believe something negative about Hillary Clinton.
Raise your hand if that surprises you. (No one should be raising their hand.)