The following is a comment posted to the publication Townhall as regards what has been the phenomenon of the left media glomming on to the recent DMR- Mediacom fostered Selzer poll supposedly indicating Trump is behind Harris in Iowa. Selzer argues to the effect that the poll cannot be considered a push for Dems because ~~ wouldn’t it make the Republicans work harder to get out the vote~~. Our argument is that the poll, if not to demonstrate incompetence, the purpose as presented is to encourage Democrats in truly battleground states.
I live in a blue county in Iowa and I can say Dems are active, more so than Repubs as far as media and yard signs. That Dems are active is not surprising, that Repub are not more is disappointing but perhaps indicative of presumptions about how the state will go. The presidential race drives too much of the spending and impinges on attention and resources for down ticket races, especially if the Dems see those races as their only hope and focus accordingly. I am looking through the blue hue here so hopefully it is rosier in most of the rest of the state.
Going by the Emerson Poll it is rosier elsewhere but Selzer is maybe loading her calls in blue counties (the 4 or 5 most populated) and all the cat ladies elsewhere in the state who are more willing to answer the phone. But I am still a bit worried about retrenchment for lack of enough fight by Republicans, congressional candidates excepted. However in more sober or less paranoid moments that the state can be predicted to flip in spite of all the other indications of Republican structural increases seems unreasonable.
If I am correctly informed of Selzer’s approach which is that she relies less on weighting and more on random calling than other pollsters, which some might argue takes pollster manipulation out of the results, she may have established the new text book example of the modern day problem with her approach, present for a number of years but this year in her rendition unameliorated by luck of the draw. More broadly referred to as the no-response bias. The problem is that “random” people do not answer their calls. A certain even if undefined “type” do and do not.
More and more “average” people do not answer their phones from unrecognized numbers and those that do might have a bias compared to those that do not. Further there is a particularly palpable resistance by Republicans and particularly Republican men to trust or respond or sit still for any political call interrogations including from honest pollsters who first must cut through the push-polling conducted by campaigns and list mining operations in order to get an interview.
And earnest pollsters are not free from sin and that is also reflected in peoples response to their calls.The tainting of their questions particularly when it comes to issues, the presumptions in the use of language is more and more understood and rejected even if not argued at the time. Their questions are stiff and fixed, and frustrating resulting in no responses, hangups, and incompletes. If more and more people do not trust pollsters, those that due may overrepresent a political nuance or view thus skewing the results. Too many pollsters may be polling unrepresentative trusting people, people who take such calls, people more likely to trust government and other people.
Perhaps Selzer and the DMR — will be properly embarrassed by their incompetence or purposeful thumb on the scale including HOW gleefully they reported the biased news. In 1948, more innocently, pollsters relied on phone contacts which, insufficiently weighted, favored Republicans. It was easy, more Republicans had non-party lines. Maybe after a solid win in Iowa by Trump — more towards the Emerson poll showing 10% margin for Trump, Trump will be able to hold up the Des Moines Register headline after he wins the election, like Truman did in the now iconic photo from 1948.
Even at that, if the DMR was honest or innocent or competent, it would only properly say the race appears to have closed, not that Harris leads, but that would not be good enough to motivate the un-voted Democrats in the other states, that hope they decided to be in service to, the true target of the headlines. It was a get out the vote effort not in Iowa, a small Electoral College state they knew was going Trump despite all their efforts over the years. It was meant for the battleground states.
The true meaning of the poll could just as well be written as Trump may get 47% to Harris 43% in Iowa as that would still be in the margin of error for both. In other words it is just as likely Trump gets 47% not 44% and Harris gets 44% not 47% and still be within the 95% confidence level from the sampling’s 3.4% margin of error.
The same confidence level and margin of error for the Emerson poll puts Trump’s support at +10 thus putting the chances beyond Harris’ reach, for that sample anyway, and assuming neither poll is a flier. One of the two polls is incredible, given all manner of other indications and trends, is significantly biased in its sample methodologies (innocent or not, ) and that poll appears to be Selzer’s.
And by the way wouldn’t an objective news organization give some analysis to an alternative poll unless they were vested (the liberal DMR is invested) in the results of the Selzer poll which they commissioned? Questions arise, did the DMR commission the results or did Selzer play to her own politics, or did she take leave of her senses in constructing and allowing publication without at least insisting on a proper presentation, if nothing else to protect her reputation (however overdone that is as she has had some doozy errors over the years)?
Proper unbiased reporting by other outlets (the usual suspects the DMR was feeding and that ate this up) is not properly a matter of just regurgitating the alleged numbers but rather to at least engage in a cursory examination of the true meaning of the numbers and avoid bias in their presentation of them, which obviously the DMR was not interested in doing. Other consumers of this crap poll if honest would at least mention the existence of a competing poll by a reputable organization, just as recent, with same sample size in same state, same margin of error and degree of confidence. At the very least the use of the Selzer poll was to push.
All that said, I am not as convinced as some of what the degree of early voting means for Republicans compared to alternative uses of resources. The time and resources expended by the formal party involvements are GOTV oriented. They do little on messaging which I maintain, if started early, done well and consistently as to why to vote R, inoculates and changes voting patterns and would be more additive than the net results the paid and volunteer drones produce by pushing early voting.
Chronically there is an inordinate expensive effort at vote by mail rather than, if it is so important to vote early, to do so more securely in person at satellite early voting locations that are reasonably generous across the state. The relative cumbersomeness of vote by mail; that it is not a good government approach to elections as it encourages, certainly enables onesy-twosy fraud; that it prevents proper voter ID, combine to create an aura against Republican interests. The costs associated with promoting one mechanism of voting rather than motivational messaging to increase support and turn more Democrat leaners into Republican voters thus turning the Dem mules into our mules — is not required for Republicans to win, biggly.
