Five years ago tomorrow, I walked into The Daily Item building in Sunbury for the first time.
I’m not really sure why, but I thought about that quite a bit last week, and about how much has changed since.
I’ve always had a thing for anniversaries. Did you know we’re marking 400 years of the landing of the Mayflower in Massachusetts this fall? You do now, and we’ll have a special page about that bit of history this week.
Nice as it is to occasionally look back, looking ahead is pretty much the full-time job in this and many businesses.
That’s not much fun right now if you happen to be a pollster.
Last week in this space, I wrote that it was highly unlikely that we would know the winner of the presidential election by the end of Election Day.
That makes me 1-for-1 in accurately forecasting something about this election. There are not many other perfect records on this topic.
W. Joseph Campbell, Professor of Communication Studies at the American University School of Communication, wrote a piece this week for “The Conversation,” a journalism partner of ours, that summed the state of polling up rather well.
The failure to get things right, he wrote, “deepened the embarrassment for a field that has suffered through — but has survived — a variety of lapses and surprises since the mid-1930s.”
Campbell also cited a line from Politico’s “Playbook” newsletter that took things further.
“The polling industry is a wreck,” it declared, “and should be blown up.”
That’s probably too harsh. To be fair, polls aren’t predictions. They are surveys that try to assess what people are thinking. There is always a stated statistical “margin for error,” — usually plus or minus 3 percentage points. And guess what? Sometimes people are less than fully honest about how they respond.
Still, considering the lessons in hand from the 2016 election, the polling industry appears to have gotten more wrong than right again this year.
Chris Ellis, associate professor of political science, Bucknell University, said as much in a chat with Daily Item reporter Rick Dandes on Thursday.
“The polling industry is definitely in for some soul-searching after this year,” Ellis said.
People tend to like reading polls — at least when they show their candidates ahead.
Here, while we run short stories on several national polls, we tend to pay the most attention to the one that comes out of Franklin & Marshall College in Lancaster. Dandes routinely speaks with that poll’s director, G. Terry Madonna. Their final poll on Oct. 29 had Biden up by 6 points.
“Trump’s supporters really came out strong,” Madonna said Thursday. How far his and other polls were really off, he added, will have to wait until the mail-ins are counted.
Maybe so. But the undeniable fact is, as Bucknell’s Ellis said, polling is “harder now than it used to be.”
Most people with caller ID don’t answer when they see unfamiliar numbers pop up on their screens. There are way fewer landlines and even fewer people willing to respond honestly — or at all — to surveys.
Moving forward, there’s going to be a lot of conversation nationwide about whether polls remain a viable election reporting tool.
We’ll do that here, too. Feel free to weigh in.