Inside the Propaganda Machine
The most important thing for voters to understand about the coming midterm elections is that almost everything you read or hear about them is not news — it’s propaganda.
The New York Times, The Washington Post, NBC, ABC, CBS, The Hill, Politico and scores of other news outlets do not offer news except on mundane or purely local issues. When it comes to key political issues, they are propagandists par excellence.
Propaganda involves taking some facts to create a veneer of credibility and bending them into narratives that are false but serve the purposes of the publisher.
This propaganda effort is not conducted on an ad hoc basis. It is coordinated through talking points that quickly spread among the participants.
There are three keys to successful propaganda. The first is repetition. It doesn’t matter how weak the narrative is. If you repeat it over and over, day after day, it will gain credibility from repetition alone.
The second is coordination. If you see a story in one publication but then see the opposite perspective in another, you might assume that there’s a fair debate on the subject matter. But if you see one version in 10 different publications, the likelihood that you assume “it must be true” because it’s ubiquitous goes up significantly.
The third element is bootstrapping. This means one outlet will publish a skewed or manipulated “story.” Other publications will immediately use that story as the “proof” for their own stories on the same subject.
Again, a cascade is created practically from thin air. Only when you know how to see through this process can you come close to getting any real news.
Don’t Take Polls at Face Value
One of the sources that feeds the propaganda machine is polling.
For example, an August Susquehanna poll showed Democrat Abdul El-Sayed ahead of Republican Mike Rogers in the Michigan Senate race by seven points. Recent polling also shows Democrat Sherrod Brown ahead of Republican Jon Husted in the Ohio Senate race. Democrats lead Republicans in other Senate race polls in New Hampshire, Texas, Georgia and other key states.
The Democrats need to “flip” four Senate seats from Republican to Democrat in order to take control of the Senate. According to these and other polls, the Democrats are almost there.
The problem is that these polls can be badly skewed in favor of Democrats. Once you adjust for the skew, you get very different results showing the Republicans are in decent shape to keep control of the Senate.
The first variable to review is the sample size. A poll that has 1,300 or more respondents can be fairly accurate, with a margin of error of roughly 3% or less, depending on the methodology. As the sample size decreases, the margin of error generally increases.
In some cases, you’ll see a sample size of only 1,000 or perhaps as low as 800. Those polls can have margins of error in the 4% range. That matters in a world where many elections are decided by even narrower margins.
Pollsters know this, but what they want are headlines, not accuracy.
Even the Skew is Skewed!
The bias gets worse. Within whatever sample you are using, you can skew the results by oversampling Democrats.
For example, think of the general voting population as roughly 40% Republican, 40% Democrat and 20% independent. The exact percentages vary depending on the survey, how party identification is measured and whether you are looking at all adults, registered voters or likely voters.
Independent voters usually determine close elections because they can lean in either direction and are more difficult to predict than mainstream Republicans or Democrats. An accurate poll should weight its sample to reflect the voters who are likely to turn out.
Instead, polling firms can construct polls that have 50% Democrats, 35% Republicans and 15% independents. Pollsters should correct for significant partisan imbalances when weighting the final sample. If that imbalance is left uncorrected, increasing the Democratic share of the sample will obviously increase the Democratic share of the result.
This is how you can get polls showing Democrats with significant leads even though the actual electorate may be more favorable to Republicans.
Then pollsters can go further and create a skew-within-the-skew. First, they can oversample Democrats. Then, within the Democratic cohort, they can oversample Black voters.
The Black segment of the U.S. population is about 12%. But some polls could make Black voters 20% of the sample. Pollsters should account for that difference when weighting the final sample.
If left uncorrected, the arithmetic matters. Black voters routinely vote heavily Democratic. When you sample eight percentage points of “extra” Black voters and they vote 90% Democratic versus, say, 50% for the general population, you are adding 3.2 percentage points to the Democratic total.
This can also be done with women voters, who lean Democrat, or with voters from particular states, including California or Massachusetts, who skew progressive.
At the very extreme, duplicitous pollsters can do all of the above: They can use a small sample size, overweight Democrats in general, then overweight Black voters, women and progressives in particular.
Then the propaganda machine kicks in. The skewed results become headline news to mobilize Democrats, discourage Republicans and suppress turnout by convincing Republicans they’re going to lose anyway.
It’s a slick method of voter suppression. And it’s critical for Republican voters in particular to look past the headlines and understand what’s really behind the polling numbers, especially if they want any chance of holding the House.
All the best,
Jim Rickards
Editor, Strategic Intelligence