Preparing for a Long Post-Election Count
July 1st 2020
CNN coverage of DNC (Matthew Reichbach)
By Walter Shapiro
Election night is at the center of the opening chapter in The Making of the President 1960, the most influential campaign book ever written. In it, Theodore White conveys the dramatic sweep of a close election as TV reports follow the time zones from Connecticut (safe for John F. Kennedy) to California (a tossup in Richard Nixon’s home state). The chapter ends with JFK nervously pacing the lawn at Hyannisport at 4:20 in the morning unable to “tell whether he had won or lost.”
Election nights have often been fraught with drama. As a long-ago junior campaign staffer, I still recall the thrill at 3:30 in the morning in 1976 when the networks called Mississippi (yes, Mississippi) for Jimmy Carter to give him an Electoral College majority. And anyone who stayed up late to watch the returns in 2000 remembers that stunning moment when the TV networks rescinded their 2:20 a.m. call of Florida for George W. Bush and labeled the state “too close to call.” As a result, it took a Supreme Court decision to determine the outcome of that race.
There was a hunger for election night returns long before the arrival of television or even radio. On the night of President Woodrow Wilson’s narrow 1916 reelection victory, for example, the New York World described the mood along Broadway: “Election returns were served with the eats; they were handed out with the drinks. They were part of the programme at the theatres. Returns were coming from everywhere.”
For well over a century, election night has been a pageant of democracy. Since the days of anchors like Chet Huntley and David Brinkley, TV broadcasts — with their red-and-blue maps, their exit polls, and their ever-changing numbers — are how Democrats and Republicans, winners and losers alike, have learned and accepted the results of a presidential election.
Until now.
Unless the 2020 presidential election is a blowout, the pandemic will likely play havoc with our election night rituals. The explosive growth of absentee ballots and their slow count will mean days or weeks of uncertainty if the election is as close as 2016, when six states (Michigan, New Hampshire, Wisconsin, Pennsylvania, Florida, and Minnesota) were decided by less than 2 percentage points. Long lines, malfunctioning machines, and attempts at voter suppression are also apt to add to the controversy and confusion of a Covid-19 election. As a recent Washington Post headline put it, "Day-of results unlikely Nov. 3."
While there are many worthy proposals to limit the damage to democracy from a slow-reporting and contentious election, almost all of them focus on procedures for voting and citizen education. But, in many ways, everything depends on how the TV networks and other major news outlets handle election night. In these troubled times, any uncertain outcome is apt to spawn conspiracy theories. Already, President Trump is anticipating election night with evidence-free tweets like this one from Monday: “Because of MAIL-IN BALLOTS, 2020 will be the most RIGGED Election in our nations history — unless this stupidity is ended.”
The networks and the Associated Press have long been meticulous about not calling states on election night until they are certain of the outcome, based on both exit polls and actual returns. In a memorable moment from 2012, Fox News anchor Megyn Kelly dramatically walked down a long corridor to the network’s decision desk after Karl Rove challenged on air the fact that President Obama had been named as the winner in Ohio. Despite Rove’s obstinate denial of reality, Obama did, in fact, carry Ohio by 165,000 votes.
At the same time, TV networks hate protracted uncertainty. Think of how rare it is to hear a television pundit say the near-forbidden words, “I don’t know.” During my lone stint as an election night commentator for ABC News online in 2012, I realized how tempting it is to get ahead of the story and offer a verdict based on incomplete returns. Observations such as “It’s looking like a good night for Candidate X” or “From what we know right now, I would certainly prefer to be Candidate Y” can inadvertently create misleading narratives.
Interpreting partial results in 2020 will be tricky. Exit polls are geared to in-person voting, although they have been tweaked in recent years because of the shift to early and mail balloting. Swing states like Pennsylvania lack a tradition of heavy absentee voting, so it will be difficult to figure out whether Democrats or Republicans are disproportionately voting by mail.
The danger lies in the likelihood that the final results in some hotly contested states will prove significantly different than the numbers prominently displayed on election night. Academics have documented in elections dating back to 1948 the so-called “blue shift” in which late-arriving absentee ballots and provisional ballots tend to favor Democrats. But Covid-19 may alter these historical patterns since different demographic groups may be voting by mail.
On election night, the TV networks are almost certain to add cautionary notes to their coverage about how vote totals are preliminary and subject to change with the arrival of mailed-in ballots. But these network public service announcements, well-intentioned as they may be, are apt to get lost amid the excitement of election night coverage.
The most important thing that the networks can do to combat conspiracy theories is to change how they display election night returns. Traditionally, the networks show the gyrating numbers as a percentage of precincts reporting. But that approach has proven inadvertently deceptive since precincts in some states only report in-person votes and others exclude late-arriving absentee ballots. But in their coverage of last Tuesday’s primary elections — despite the massive shift to absentee ballots thanks to Covid-19 — the New York Times maintained the ludicrous fiction that 50 percent of the precincts in the closely-watched Kentucky Democrat Senate primary had reported, even though this count only included in-person ballots.
Instead, the networks should display the returns as a percentage of the estimated turnout. Using polls, absentee ballot requests, and actual returns, the networks are sophisticated enough to come reasonably close on election night in predicting the total number of votes cast in each state. It would make a vast difference if voters know that even though a presidential candidate is ahead by, say, 50,000 votes in Wisconsin at 2:00 in the morning, an estimated 38 percent of the ballots in the state remain uncounted.
In normal times, television networks should not be expected to be the final bulwark guaranteeing that both sides accept elections as legitimate. But the arrival of a pandemic at the height of partisan discord has upended traditional rules. As a result, the election night 2020 broadcasts may be the most fateful for democracy since Philo Farnsworth demonstrated his first television set in 1927.
The views expressed are the author’s own and not necessarily those of the Brennan Center.
Posted with permission