For Polling, It’s All About the Questions

Commentary

This past December, a poll began to circulate around Washington that caused a sudden shift in messaging from the White House and beyond. It had a stark conclusion: “Republican and Democratic candidates who support eliminating long-standing vaccine requirements will pay a price in the election. … Vaccine skepticism is bad politics.”

Sheryl Gay Stolberg of the New York Times identified this poll as the one that rattled Republicans. “Just one in five voters approves of rolling back established vaccine recommendations,” she summarized. As a result, the political environment on the topic suddenly changed. The vaccine industry is off the hook.

Curious, I took a closer look at the poll. It turns out to be constructed to come up with precisely that conclusion. I will explain more below, but consider that every poll has this problem. Anyone can get answers they want provided that the questions are framed in a certain way.

That’s not to say that people don’t have real opinions on matters. These opinions are worth discovering. The trouble traces to the interfacing between the poll creator and the polled person.

In the most famous case, support for Trump as president has been consistently underestimated for 10 years. The reason is that people are reluctant on the margin to admit their views, knowing full well that the mainstream media cannot stand the guy. So the polled person often takes the easy route.

This is called the social desirability bias. This is a polling/survey error in which respondents give answers that they think are more socially acceptable or approved of, rather than their true views. That means the results end up underreporting undesirable behaviors or opinions (support for a polarizing figure) and overreporting desirable ones.

A pollster with an agenda can use this bias. Let’s say you want a poll to reveal vast support for carbon taxes and the Green New Deal. You could field a question such as this one: “Do you agree with the scientific consensus that climate change is both real and a problem?” What do you think the results would be? I’m guessing that 75 percent would agree.

But you could put it another way: “Do you favor new taxes on fuel and other controls to curb carbon in the hopes that this will change the climate in 20 years?” I’m guessing that the results would be the opposite. The difference is that the first question encourages people to say yes and the second one invites them to say what they really think.

Again, it’s not that people have opinions that blow with the wind. It’s that being polled is inherently intimidating. Even tiny changes in wording can make a big difference in the outcome. Polling companies know this and respond to their clients. If “Citizens Against Climate Change” funds the poll, it is going to turn out to favor their agenda.

There are plenty of other biases, such as the type of people willing to be polled at all. A few years ago, for example, a polling company called me to ask my opinions on various national matters. I was willing to answer questions. But even before the survey began, I fielded questions about my race, age, income, living arrangements, job, and so on. I was already on the hot seat. I hung up simply because I’m unwilling to share with a stranger personal details that I wouldn’t even tell a cousin.

With all of this in mind, let’s now look at this poll that claims to show that vaccine skepticism is a political loser. It begins with a statement and asks whether you agree or disagree. The statement is: “Vaccines save lives.” Can you imagine the results? Among Kamala Harris voters, it is 99 percent. Among Trump voters, it is 75 percent. Among swing voters, it is 83 percent.

The question itself is preposterous. To answer no to it, you would have to believe that in the 228 years since the invention of vaccines, no lives have been saved. Not from smallpox and not from rabies. That’s completely ridiculous. It is rather obvious that vaccines have saved lives. It’s not even in dispute. That a quarter of Trump voters said no is actually astonishing.

A second statement reads: “Vaccines are the best defense against many infectious diseases.” Once again, it is nearly impossible to disagree with that statement. Sure enough, 70 percent to 90 percent of people are compelled to agree.

For sterilizing vaccines such as measles or chickenpox, they do indeed provide the best defense against getting the disease. The statement does not address whether getting the infection when you are young is a better approach for gaining lasting and broad immunity—which is hardly disputed either. But the statement does not allow for that subtle distinction.

In other words, this part of the poll is wholly invalid.

Now to the next part, which attempts to measure how important issues of vaccines are to voters.

The poll asks people to rank the importance of these issues. Notice which ones are easy, which ones are harder, and which of the eight lends itself to a solid “not interested.”

1. Require labeling of harmful ingredients and chemicals in ultra-processed foods
2. Remove toxic chemicals and pesticides in agricultural practices
3. Remove from foods and beverages dyes and artificial ingredients, such as red dye 40 and high fructose corn syrup
4. Restrict federal subsidies for ultra-processed foods and beverages in school lunches
5. Remove genetically modified organism ingredients from our food
6. Prohibit direct-to-consumer drug advertising
7. Stop adding fluoride to local drinking water
8. Remove established childhood vaccine recommendations for diseases such as whooping cough, measles, hepatitis, and others

You don’t need to be told that question No. 8 came in dead last, which allows the poll to be interpreted as proving that changing the schedule is politically unpopular.

Let’s just take apart the sentence as it reads: “Remove established childhood vaccine recommendations for diseases like whooping cough, measles, hepatitis and others.”

It’s the longest and most complicated of all. The words are complicated and strung together like linguini.

Note the term “established.” It means settled, done, recognized, practiced, fixed in place. To oppose what is established is to upset the apple cart, and you need to have good reasons for doing so. Also note the use of the term “recommendations” rather than requirements, which is what they are in practice. Who can oppose an established recommendation?

Also note that the recommendations are “diseases like” three terrible things. If there is a way to stop those things using established recommendations, who in his right mind would oppose that?

This is how you build a biased poll, not with obvious language but with little sprinkles that trigger the mind in a certain direction. It’s extremely clever.

Remember that most pollsters work for clients. Their job is to make the person paying the bills happy. This is why it is absolutely essential to look under the hood, so to speak, and examine the questions from the point of view of a normal person. Above all else, don’t let your own views on anything be affected by a poll that may or may not be valid. It could just reflect the bias of the poll’s designers.

Keep your decoder ring on. Surveys and polls have long been used to manipulate the public mind. You don’t want to be among the manipulated.

Views expressed in this article are the opinions of the author and do not necessarily reflect the views of The Epoch Times.