Was Cable TV Election Poll Coverage a Waste of Time?
Public polls are generally ill-equipped to fully capture changes happening in the last days of the campaign.
NEW YORK (AP) — It's understandable if Bill Hemmer, John King and Steve Kornacki still see flashing maps of blue and red states in their mind's eye before drifting off to sleep.
Each man was assigned by his television network to stand before a map of the U.S. several times a day during the presidential election campaign to talk about the latest polls and speculate on "paths to victory" for Hillary Clinton and Donald Trump.
Given the Republican nominee's surprise win, it's hard not to wonder whether it was wasted time that in future campaigns might be better spent trying to understand voters or the more substantive issues facing a new president.
"I do not feel it was a wasted political exercise at all," said Fox News Channel's Hemmer. "Based on the level of national interest in this story, people were hungry for information and it was our duty to provide that."
The one dominant theme of the map-side discussions was that Clinton had the clear advantage, and that many things had to go right for Trump to win. "If you Googled it, you would probably hear the phrase 'inside straight' several times, because that was what they needed," said Hemmer.
There were signs in the campaign's final days that things were tightening, and it was reflected in the reporting. Nate Silver of ESPN's 538 blog, in fact, was sharply criticized by Clinton supporters the weekend before the election for not being as bullish about their candidate's chances as others were.
Even on the afternoon of Election Day, pollster Ed Rollins said on Fox it would take a miracle for Trump to win. The New York Times' Upshot blog, which carried a constantly updating dial on each candidate's chance of winning, early that day pegged Clinton's chances at 84 percent.
Forecasters like Silver, who built his reputation on his 2012 success, increased the appetite for Hemmer, King and Kornacki's work. Obsessives hung on every word, every poll.
"I was as shocked as everyone as it turned on Election Night," said MSNBC's Kornacki. "There was about a 20-minute period, looking at Florida, in North Carolina and in Virginia, that just turned upside down everything I thought about where this was heading."
Kornacki takes some comfort in the knowledge that he repeatedly told viewers that Rust Belt states like Michigan, Pennsylvania and Wisconsin were the secret to any Trump victory scenario.
Four days before the election, Trump's polling team came to Fox News to show Hemmer that their research was more positive than public surveys were reflecting. They turned out to be right; at the time, Hemmer had to worry if he was being spun.
Polling captures a moment in time when only one moment — Election Day — really matters. And the public polls are generally ill-equipped to fully capture changes happening in the last days of the campaign.
Thomas Patterson, a professor who teaches about politics and the press at Harvard University's Kennedy School of Government, wrote a decade ago about the flood of polls damaging campaign coverage. He called them a cheap branding opportunity for universities and media organizations.
Things haven't changed since then, he said.
"The problem with polls when you have that many is that it's cheap and easy journalism," Patterson said. "The audience is interested and you get new stuff every day. At some point, when you pay that much attention [to polls], it hijacks the news."
In an upcoming study, Patterson found that no substantive issue brought up by Clinton attracted more than 1 percent of campaign coverage. The candidates shoulder some blame for the type of campaigns they ran, but the finding also reflected how it was covered, he said.
Polls are especially tempting for cable networks with endless hours to fill, where talking about the news often beats reporting it. They become the mirror that reflects everything; much of the discussion over Clinton's use of a private email server was not about the issue itself, but rather how it would affect her popularity, Patterson said. Clinton was often covered as the next president, Trump as a curiosity and sure loser.
Michigan-based filmmaker Michael Moore, who predicted a Trump victory last summer, is seen now as a seer for his knowledge of a region often overlooked by national news organizations. His post-election appearance on MSNBC's Morning Joe was gripping television, and the failure to read attitudes in the Midwest has dominated media post-mortems.
CNN, which declined to make King available for this story, hired as a contributor Salena Zito, a former Pittsburgh newspaper reporter who, like Moore, had seen clearly what was coming.
MSNBC's Kornacki, a political junkie who volunteered for his map duty, argued that studying the polls is as important to spot trends as better regional reporting, even with the spotty performance this year. "We cannot go out and conduct 350 million interviews across the country," he said.
The veteran political observer Patterson needs no poll to predict what the current media self-examination will result in.
"When the flag drops the next time," he said, "I don't think it will be that different."