New York Post finds a big problem with that Trump impeachment poll

I think it was Megan McArdle, or someone like her, who noted that the important thing about reading news these days is not the news itself, but who wants us to know.

Which is what makes push polls very attractive vehicles.

The New York Post smelled a rat with a recent Fox News poll, which then drew huge headlines from apparently Trump-hostile billboards such as the Drudge Report, and decided to do a deep dive:

The poll released last week by Fox News that claimed most Americans favor the impeachment of President Trump underrepresented Republican and independent voters, The Post has found.

The poll said 51% of voters were in favor of Trump’s impeachment and removal from office, while 40% did not want him impeached.

Princeton, New Jersey, pollster Braun Research, which conducted the survey, noted 48% of its respondents were Democrats. But the actual breakdown of party-affiliation is 31% Democrat, 29% Republican and 38% independent, according to Gallup.

The Fox poll oversampled registered Democrats, well out of proportion to their true numbers and then tried to pass its result off as consistent with public sentiment. Funny how such mistakes never happen with polls overweighting Republicans.

Update: Fox News defended its poll, however. On Sunday, a representative of Fox News Media Relations provided the following statement by FOX News Vice President of Public Opinion Research Dana Blanton:

“Our polling unit has long been held in high regard for being a nonpartisan source of research. Under the joint direction of Beacon Research (D) and Shaw & Company (R), the latest FNC poll included interviews with randomly chosen registered voters and – as is our standard practice – we reported the partisan distribution we found among the electorate. Braun Research is solely our data collection partner. We stand by our latest poll.”

O.K., fine. I still don't understand why the poll oversampled Democrats, as the Post argued.

Pollsters can correct for such discrepancies without throwing out any data, through math, and the Post wasn't shy about trying just that to show a more professional and reliable count.

The Post found that if the nunbers were corrected for voter proportionality, then 44.9% of voters favored impeachment and 44.2% didn't. That's still not good, but it's a whole lot less on the spread count than the 51%-40% figures reported in the Fox News/Braun Research poll. So color that debunked and take a look at who was behind it and wanted us to think it.

The Post said the culprit was Braun Research of Princeton, New Jersey, and they are the ones who presumably didn't do their math.

Who's the biggest fan of impeachment among the Democrats? All of them, of course, which fails to narrow the field, but one candidate is Tom Steyer, the megabucks greenie billionaire who last year spent millions to run anti-Trump impeachment ads, despite a lack of public support, which the ads did nothing to help.

Does he have any connections to Braun? Not directly so far as I can tell (there are still plenty of places to look) but recall that Steyer last got his name in the news for mysteriously qualifying for the fourth Democratic debate in November, despite no obvious well of support. There were several criteria to qualify but one of the qualification rules required breaking 2% in approved polls, one of which was Monmouth. This may be putting two and two together and getting 22, there could be contexts that erase this as a factor, but one of the research firms that does work for Monmouth ... is Braun Research.

Nice coincidence, if nothing else. 

Someone out there wants Trump impeached and wants Democrats to be emboldened enough to pursue it based on a reduced risk of political repercussions. Is Steyer the kind of guy who might have money for certain kinds of polls getting out? He is. He's not the only one but my antennae are up. 

Photo illustration by Monica Showalter with Pixabay public domain image

 

 

 

I think it was Megan McArdle, or someone like her, who noted that the important thing about reading news these days is not the news itself, but who wants us to know.

Which is what makes push polls very attractive vehicles.

The New York Post smelled a rat with a recent Fox News poll, which then drew huge headlines from apparently Trump-hostile billboards such as the Drudge Report, and decided to do a deep dive:

The poll released last week by Fox News that claimed most Americans favor the impeachment of President Trump underrepresented Republican and independent voters, The Post has found.

The poll said 51% of voters were in favor of Trump’s impeachment and removal from office, while 40% did not want him impeached.

Princeton, New Jersey, pollster Braun Research, which conducted the survey, noted 48% of its respondents were Democrats. But the actual breakdown of party-affiliation is 31% Democrat, 29% Republican and 38% independent, according to Gallup.

The Fox poll oversampled registered Democrats, well out of proportion to their true numbers and then tried to pass its result off as consistent with public sentiment. Funny how such mistakes never happen with polls overweighting Republicans.

Update: Fox News defended its poll, however. On Sunday, a representative of Fox News Media Relations provided the following statement by FOX News Vice President of Public Opinion Research Dana Blanton:

“Our polling unit has long been held in high regard for being a nonpartisan source of research. Under the joint direction of Beacon Research (D) and Shaw & Company (R), the latest FNC poll included interviews with randomly chosen registered voters and – as is our standard practice – we reported the partisan distribution we found among the electorate. Braun Research is solely our data collection partner. We stand by our latest poll.”

O.K., fine. I still don't understand why the poll oversampled Democrats, as the Post argued.

Pollsters can correct for such discrepancies without throwing out any data, through math, and the Post wasn't shy about trying just that to show a more professional and reliable count.

The Post found that if the nunbers were corrected for voter proportionality, then 44.9% of voters favored impeachment and 44.2% didn't. That's still not good, but it's a whole lot less on the spread count than the 51%-40% figures reported in the Fox News/Braun Research poll. So color that debunked and take a look at who was behind it and wanted us to think it.

The Post said the culprit was Braun Research of Princeton, New Jersey, and they are the ones who presumably didn't do their math.

Who's the biggest fan of impeachment among the Democrats? All of them, of course, which fails to narrow the field, but one candidate is Tom Steyer, the megabucks greenie billionaire who last year spent millions to run anti-Trump impeachment ads, despite a lack of public support, which the ads did nothing to help.

Does he have any connections to Braun? Not directly so far as I can tell (there are still plenty of places to look) but recall that Steyer last got his name in the news for mysteriously qualifying for the fourth Democratic debate in November, despite no obvious well of support. There were several criteria to qualify but one of the qualification rules required breaking 2% in approved polls, one of which was Monmouth. This may be putting two and two together and getting 22, there could be contexts that erase this as a factor, but one of the research firms that does work for Monmouth ... is Braun Research.

Nice coincidence, if nothing else. 

Someone out there wants Trump impeached and wants Democrats to be emboldened enough to pursue it based on a reduced risk of political repercussions. Is Steyer the kind of guy who might have money for certain kinds of polls getting out? He is. He's not the only one but my antennae are up. 

Photo illustration by Monica Showalter with Pixabay public domain image