The 2020 election continues to astound. Data, depending on the source, portray a completely different picture of what is occurring in America today. This is not normal.
Here, we take a look at the difference between polls conducted in the last 24 hours regarding ‘who won the last debate’. The one major traditional poll was conducted by phone interview by CNN and showed that Biden won with 53% support versus 39% for Trump.
The average of 10 snap polls conducted on-line by local TV or news sites states that Trump won with 66% support versus 32% for Biden.
The difference between these results is extraordinary and highly unusual. This cannot be stressed enough.
As we have mentioned many times, our alternative data does not support the Biden landslide victory scenario that is overwhelmingly prevalent among forecasters and pundits. In fact, it shows that the 2020 election should be fairly similar to that of 2016 with Trump winning the election thanks to a series of relatively narrow victories across multiple ‘swing states’.
The most important highlight from this post is that the 10 snap polls conducted by local TV stations and news sites (mostly local newspapers) supports our view.
We find it extremely interesting that the local snap polls are mostly anonymous. We have repeatedly stated that in the current contentious political environment of the US, many Trump supporters do not feel comfortable responding in a public forum. However, when anonymous information is used, there appears to be many more Trump supporters. Whether it’s an anonymous on-line snap poll after a debate (such as those highlighted here) or which sites they frequent for their news (conservative or liberal biased), people tend to be more truthful (in the current environment) when they feel there will not be recursions. In other words, the more anonymous the better.
The CNN poll, though surely conducted with a high degree of professionalism, used known respondents. And, these respondents were even paid for participating. In a normal political environment, it might not make a difference to set up a poll under these conditions. But, in today’s environment, we have our doubts.
Table 1: Polls Conducted Post-Debate, Last Presidential Debate between Trump and Biden, Oct 22nd 2020
Source: ZettaCap, WEVV, WPVI, KRQE, ActionNewsnow.com, WJW / Fox8, KOMO News, Gwinnett Daily Post, Boston Herald, NJ.com, CantonRep / The Repository, CNN
An interesting point is that many of the snap polls were taken in Democrat leaning areas such as Boston and Philadelphia. There are a number of swing states as well that are included such as Georgia, Pennsylvania, and Ohio.
The data sources are actually fairly diverse – many different regions are represented, some are site based other snap polls were on twitter, some were conducted by local TV stations and others by local newspapers.
The common theme is that the organizers were local news organizations.
Next, we take the simple average of the local TV and local news cohorts.
Table 2: Summary of Polls Conducted Post-Debate, Last Presidential Debate between Trump and Biden, Oct 22nd 2020
Source: ZettaCap, WEVV, WPVI, KRQE, ActionNewsnow.com, WJW / Fox8, KOMO News, Gwinnett Daily Post, Boston Herald, NJ.com, CantonRep / The Repository, CNN
The CNN poll tells an entirely different story. It shows that Biden will blow Trump away on Election Day.
The averages of the local news cohorts, in contrast, paint a fairly consistent picture of Trump being by far the stronger candidate.
Which is correct?
As the anonymous snap polls conducted by local news organizations more closely match our alternative data, we believe they are closer to correct.
However, the CNN post-debate poll closely matches the mountain of traditional national and swing state poll data that puts Biden well in front of Trump.
We had initially thought that these datasets would converge upon Election Day. Meaning, the large divergence would diminish, maybe not disappear, but certainly decrease to the point of being near a margin of error. This is not the case. In fact, instead, we have two very different versions of reality being painted.
On Election Day, we do not see much room for a middle ground. Either traditional polls will be proven as correct and there will be a Biden landslide or our alt data (which is supported by some other data such as these local anonymous snap polls) will be correct and Trump will win with Republicans likely retaining the Senate and winning seats in the House.