We applaud the AP for revising its guidelines to de-emphasize poll results for election coverage. The 2018 AP Stylebook will include a new chapter on polls and surveys, adding details to help journalists report responsibly on public opinion research heading into the U.S. midterm elections. The AP Stylebook forms the backbone of the standards used not just by the AP, but by the majority of news organizations around the country.
We wrote last month about how exit polls can be misleading, in fact, we made them our Measurement Menaces of the Month. Now the AP has taken decisive action, specifically rejecting the use of poll results “that seek to preview the outcome of an election” as the lead, headline, or single subject of stories.
In other words, the new guidelines are intended both to de-emphasize the horse-race aspects of election coverage and to encourage journalists to use high-quality polls. The AP has now codified what those of us in communication measurement have known forever: the mere existence of a poll is not enough to make news.
The move has been widely applauded. Former Obama White House Communications Director Dan Pfeiffer and Crooked Media contributor commented, “The media would rather use polls to tell us what it thinks is going to happen than explain what is actually happening. The AP’s decision to spend less time writing about polls is a sign that finally someone has learned a lesson from 2016.” Listen to the FiveThirtyEight Politics podcast discuss the new poll guidelines here.
Journalists take note: According to the new guidelines, the AP will accept some polls in which respondents opt in, as long as the survey includes a component in which other respondents are randomly selected. In the past, the AP rejected opt-in surveys.
Congrats to the Associated Press, our Measurement Maven of the Month, for setting higher standards for the use of polls in reporting on elections. ∞