Daily PUMA Column - Commentary by Alessandro Machi

Tuesday, November 3, 2020

Could Undersampling of Landline Users cause Inaccurate 2020 Presidential Polls?

So many polls are showing Joe Biden with a stable and commanding lead it seems next to impossible that he could lose. Is there something that the pollsters missed? 

One issue that I think could create confusion is the rather remarkable difference in the Northeast, where Biden apparently holds a 56% to 32% advantage. If California and Illinois are running 2 to 1 for Biden while National polls show Biden with "only" a 5% to 6% national lead, one might wonder if Donald Trump can eeke out several close wins in key states while losing bigly in other States. 

There apparently is another issue, Trump may have improved his percentages with both Latinos and African Americans compared to 2016.

However, there is still another issue to consider....

I think there is something that the pollsters may have missed, and that is landline polling.

A quick google search claims that only 6% of homes have landlines only. However, 40% of all homes have both landlines and smart phones. Why does this matter?

The 40% landline users could be being undersampled by pollsters, here's why. The cable companies that control a significant percentage of the landline customers in the U.S. have developed possibly "over efficient" spam detectors that weed out spam calls either before the phone rings, or after one ring. Unfortunately, the landline spam blockers also weed out other calls that could include phone numbers from polling companies. I recall seeing on my TV caller ID a phone number claiming to be from Donald Trump headquarters zapped after one ring. This happened on at least two different days.

Smart phones offer similar spam block technologies as well but for some reason I don't think they are as aggressive as the land line spam blockers. Perhaps more importantly, land line users tend to need 3 to 4 rings, even 5 to 7 rings before they get to their phone. Many times the auto callers will either give up after 3 rings, or,   grab another auto call sent out at the same time that has already been picked up, thus disconnecting the landline call for being "too slow" to answer.

So, with no verifiable proof other than my own experiences with both landlines and an old cell phone, I think it is possible that a significant percentage of land line users are not being mathematically polled regarding how they will vote versus how smart phone users will vote.

So the next question is, are landline users more likely to vote for Trump, or Biden? That is the mystery question. If Trump wins the 2020 Presidential election, then perhaps it may be attributed to the forgotten landline community that still equals 40% of all households, and of course, an electoral college system that can produce a Presidential victory even if one candidate loses by 5 to 6 million votes.


Please consider signing the Debt Neutrality Petition by clicking here.

No comments:

Best Quality VHS to Digital Transfers

Best Quality VHS to Digital Transfers
Serious Customers Welcome.

Share Gadget

10,000 Dollar Grant! Another Great Find from FABULOUSLY40.com

10,000 Dollar Grant! Another Great Find from FABULOUSLY40.com
Would this be a good way to win funds for Louisa's Law ?