Trump is touting some more fake news, as usual.
Polling firm Rasmussen recently put out a poll showing that Trump’s approval rating is about 50%. Trump, of course, got really excited about it, because it means that all the other polls showing his approval in the mid-to-low 30s are fake, and all the negative press about him isn’t actually hurting him. But the Rasmussen poll is flawed, both for not revealing its methods of sample selection, and for using a polling method that heavily favors older voters, who are much more likely to support Trump. Per CNN:
> > * Rasmussen surveys "likely voters" in their daily approval polls. They argue it gives them a more accurate reading of the actual voting electorate -- and certainly a more accurate one than polling firms that test all adults. The problem, according to Agiesta, is that we have no real idea of how Rasmussen decides who a "likely voter" is. Is it based on the 2016 election? 2012? Some combo? Rasmussen won't say. ([Here's what Rasmussen does say about its samples](http://www.rasmussenreports.com/public_content/politics/obama_administration/august_2013/obama_approval_comparing_the_numbers).) > > > > * Rasmussen used an automated voice (as opposed to a real, live person) to conduct their polls. They are, therefore, barred by law from contacting people on cell phones. People who only use cell phones tend to be younger and more diverse than those who have landlines. And younger, more diverse voters have heavily favored Democrats in recent elections. > > > > * Rasmussen doesn't disclose how they conduct the telephone portion of their surveys. (They also do an online poll as a supplement to their phone results, which is a whole other problem.) Most pollsters use a technique called [random digit dialing](https://www.icpsr.umich.edu/icpsrweb/instructors/setups/notes/sampling-telephone.jsp), which is exactly what it sounds like, to build a representative sample. Others -- particularly partisan pollsters -- poll off a privately maintained voter file. (More on that [here](http://www.slate.com/articles/news_and_politics/victory_lab/2011/12/_likely_voters_lie_why_private_campaign_polls_get_such_different_results_from_public_media_polls_.html).) Which one does Rasmussen use? We don't know if they use either -- or some other way to develop a random sample of the public -- because they don't release details on how they do it. > >
This poll is quite likely an outlier:
> > Most polling on Trump's approval ratings suggest Rasmussen's 50% number is an outlier. [Gallup's latest daily tracking poll](http://www.gallup.com/poll/201617/gallup-daily-trump-job-approval.aspx) -- current as of Friday afternoon -- puts the President's approval at 41% and his disapproval at 53%. [Marist found](http://maristpoll.marist.edu/414-international-tensions-heightened-say-many-americans-trump-approval-rating-at-39/) his approval rating at 39% and disapproval at 49%, while [CBS News pegged](http://www.cbsnews.com/news/do-americans-think-president-trump-is-doing-a-good-job/) his approval rating at 43% and disapproval at 52%. > >
And of course, even if Trump was polling around 50%, it wouldn’t make all the things he’s doing OK, it would just mean they have more support.
Stay vigilant, stay skeptical.