The phone rang, and this time, I was in no mood to get another random call after already getting a ‘your auto warranty’ and ‘your credit card rates’ offers.
On the other end of the line was a person who sounded young. He was doing a poll, and he asked me if I’d participate.
As we say sometimes in the South, I blew a gasket.
I don’t remember what I said because I was in the middle of working on a book and I was highly annoyed the phone rang. I do remember mentioning mobilization of bias and the worthlessness of polls in general. The lad hung up on me, and I don’t blame him. Hopefully my number got redlined if there is such a thing among pollsters.
I rarely participate in surveys or polls. In years past I did, and I deliberately threw them. If the surveyor asked me would I vote for candidate A or B, I chose the candidate I wouldn’t vote for. It was my way of payback for being interrupted at what always seemed to be an inconvenient moment.
I’ve made it a point to ask people I know if they participate in phone polls. Most tell me they don’t. I think my informal assessment is pretty useful because my circle is very diverse. I’d say if you handed out diverse circle awards, I’d get one. I may despise random phone calls from strangers, but I love people in all shapes, sizes, and personalities.
I’ve often had people tell me how flawed polls were in 2016 ahead of the presidential election. They really weren’t flawed, at least many of them, though, in the sense we think. Polls that do a national sample on a presidential election have never made sense to me. In the United States, we do not have a mob-ruled democracy. We have a republic. The Electoral College makes it possible for small states to have a voice in national elections, ensuring that one or two large states do not permanently select a leader for the entire country.
The losing candidate in 2016 made much of the ‘win’ on the popular vote. That popular vote was in line with national polling. The fault existed in doing national polling on a matter resting on state outcomes. I don’t care how much you ‘weight’. The US is a very diverse country and I personally believe it is almost impossible to do an accurate national poll because of our cultural differences among states.
A state poll can be useful, even in states where early voting has already occurred and results to date on turnout (not candidate or issue) are known for each party. A limited national survey of voters in different states means absolutely nothing in my opinion.
It is true that poll results can be used in an attempt to influence voters. It is false that any poll can declare, “Americans agree…” on any topic because a poll is a snapshot in time, and any identifying data is based on self-reportage.
Leading questions can be asked to deliberately sway opinion. False answers can be given for various reasons. If a surveyor asks, Did you vote in the last election?, a person might say s/he did to avoid embarrassment.
We tread dangerous ground when we set policy decisions by polls. The only poll that counts on any issue or candidate is the exercise of our right to vote on Election Day.
Harvard Business Review acknowledged the difficulty in conducting polls in today’s climate where not everyone has a landline, it’s uncertain who’s on the other end of a cell phone, and the sheer numbers required are daunting:
Before the widespread use of caller ID, telephone polls were much easier to carry out. In 1997, a national sample of about 800 respondents required between 2,000 and 2,500 calls. Today, getting that same number of respondents requires between 7,500 and 9,000 calls to get a reasonably sized sample, a precipitous decline in what’s called the response rate, which is seen as a crisis in the industry. This is a problem because of the increased cost, but also because it could mean that the sample doesn’t actually reflect the group that the poll is trying to measure.
In today’s culture of revenge, often fueled by social media, people can be wary about disclosing any personal information. We also live in the ‘Age of the Scam’, and that tamps enthusiasm for talking to strangers on the phone.
Next time you see a media outlet make a blanket declaration about all Americans and base that declaration on a poll or survey, remember that no one can accurately gauge what all Americans think anymore than your local weathercaster can always get tomorrow’s forecast right.
A poll is a snapshot in time and it’s almost always based on variables. That alone makes me unwilling to give up my information, even if it’s a simple ‘yes’ or ‘no’ when I get nothing in return. Those who are called to participate are also trusted to self-report answers on everything from voting to gun rights.
I might do a poll if someone paid me. After all, the pollsters are making plenty of dough off the folks who cough up private information.
Next time you see a pollster declare, “Americans agree…”, take it with a grain of salt because it is almost always an absolute lie if it is framed that way in the headline.
(Kay B. Day/July 21, 2020)
We struggle to continue to publish.
The new Web is weighted towards big partisan media outlets, not small publishers. The only way we can stay online is by relying on the support of our readers.