Let’s just agree to stop looking at the polls
“Yes, we are officially in the Democratic bedwetting era for the 2024 presidential election. But here’s some advice from someone who’s been here before: Don’t panic.” — Jim Messina, former Barack Obama campaign head, writing in Politico Magazine
I remember when caller ID came out.
You had to buy a little device that connected to your phone. Phones didn’t have built-in caller ID until later.
Screening phone calls suddenly became a lot easier. Every bill collector could go to the answering machine. Fighting with your sister? Let the phone ring when you see her number pop up. Not sure who’s calling? Just let it go to the answering machine to find out.
I haven’t picked up a call from an unrecognized number since we plugged in our first caller ID some 30 years ago.
I tend to think a lot of people operate the same way I do.
And I tend to think that’s why political polling has grown so unreliable in recent years.
In 2012, polls predicted big trouble for Barack Obama’s reelection. He trounced Mitt Romney.
In 2016, polls showed a relatively easy win for Hillary Clinton. She lost to Donald Trump.
In 2020, polls predicted a runaway Joe Biden win, but Biden squeaked by in a lot of states.
Now, the latest polls show Trump besting Biden in several important swing states, and Democrats have gone all aflutter, wondering if Biden needs to drastically change his campaign or even drop out of the race.
I fail to see why Democrats have worked themselves into such a tizzy.
Not that Biden has no problems.
While inflation has cooled for months on end now, almost everything costs a lot more now than it did when Biden took office, dragging down public perception of the economy. Biden’s embrace of the term “Bidenomics” just doesn’t seem to be working.
Many — especially those in the most progressive wings of his party — disagree with Biden’s full-throated support of Israel in its war against Hamas in Gaza.
And a lot of people worry about Biden’s age and question whether he’s mentally and physically fit enough to work in the West Wing for another four years.
All of which could be reason enough for some people to want Biden to change course or drop out.
But I don’t think the polls should worry anyone.
Here’s my theory as to why:
Nowadays, the caller IDs on everyone’s smartphone alert us when someone calls from an unrecognized number, and a lot of people — especially a lot of younger people — let such calls go straight to voicemail. That means pollsters end up leaving a lot of messages, and I doubt very many people call them back.
Now, pollsters will tell you they account for such aggravations by simply calling more numbers until they’ve contacted a large enough sample of whatever group they’re trying to poll.
But I think the straight-to-voicemail crowd represents a particular group of people who affect elections but haven’t been captured by pollsters in years.
Maybe they’re people who vote on Election Day but don’t pay close attention to politics otherwise. Maybe they’re busy working individuals who simply don’t have time to answer unknown callers but still care about politics. Maybe they’re people who trust their candidate but don’t trust pollsters.
Whoever they are, they obviously aren’t speaking to a lot of pollsters, meaning they’re not represented in the poll results.
But they obviously vote, swinging elections in a direction different from the direction implied by people who do talk to pollsters.
The absence of that group (or groups) of people from the poll results makes those results less accurate.
Yet marketing research and public opinion polling remains a $21 billion industry as political parties, newspapers, political action committees, and others pay pollsters to try to find out how people plan to vote.
The results of such polling still shift the political landscape. Political donors move money behind people doing well in the polls. Only people polling at a certain level can get on debate stages. People polling poorly drop out of the race.
That keeps happening despite a decade’s worth of poor performance by pollsters.
Until the pollsters can figure out how to reach that crowd of caller ID-watchers, I say we all just agree to stop looking at polls as anything more than entertainment. They’re a cute way to know who’s faring well among those who answer the pollsters’ calls, but not a good way to predict who might actually win an election.
Justin A. Hinkley can be reached at 989-354-3112 or firstname.lastname@example.org. Follow him on Twitter @JustinHinkley.