Projecting races before polls close is downright scary
© Getty Images

There is a new player in the game of predicting election results and it is awfully eager to shake up the system and take on the big networks. Usually, I would welcome and celebrate this novel enterprise, but it strikes me as being downright scary and disruptive, one that would alter the way election results are reported and change voter behavior on Election Day.

ADVERTISEMENT

VoteCastr is the name of this new enterprise. It plans to reveal real-time projections of the Nov. 8 presidential and U.S. Senate races in battleground states (Florida, Ohio, Pennsylvania, Colorado, New Hampshire and Wisconsin), although it won't actually call elections. Its partner is the well-respected online outlet Slate.

Projections will be made on the basis of those turning out to vote on Election Day, through a process called "predictive turnout modeling." That's a fancy term for determining if your candidate is winning or losing by selecting various voting precincts and then predicting results by assessing how many of what should be your voters are actually casting votes.

Ken Smuckler is the founder and originator of VoteCastr. (A personal disclosure: I have known Ken for nearly 30 years and respect and like him and consider him a good friend, but my affection stops in regard to this professional endeavor.)

Smuckler attempts to lessen the potential damage by saying in a Sept. 12 New York Times article, "It's what campaigns do." He then went on to say, "We're flipping up the kimono and letting people see what campaigns do on Election Day." Great quote, Kenny. But it masks the real objective: VoteCastr is a for-profit enterprise employing Silicon Valley technicians who will gather this data and then sell it to anybody who wants to buy it. In the field of voter analytics, this is called "secondary data."

All this information will be published constantly throughout Election Day. Smuckler, in an interview, described it to me in benign terms. They are acting "like a sportscaster," "just showing people the score as the game is being played." And to sum it up and calm any fears? "We're doing play-by-play."

That fear is based on, in Smuckler's words, "one night in 1980." That was the election night when President Carter conceded early, before the polls on the West Coast had closed, based on network projections. Two prominent Democratic congressmen, Al Ullman of Oregon and Jim Corman of California, later blamed their defeat on the network's actions. Their thesis was that people saw and heard the election projection and Carter's concession speech and did one of two things: decided not to vote at all or worse, while standing in line waiting to vote, proceeded to stop waiting and went home.

Smuckler calls this all anecdotal, and here is the clincher of his argument: that 1980 "created the myth of evilness of Election Day data." And then he assured me their purpose is "not to call the outcome."

Julia Turner, editor-in-chief of Slate, seeks to take the high journalistic road. She stated in an interview that their partnership was driven by the desire "to give more information" and "give better information." She started the interview by proudly saying that Slate had a "history of leaking exit polls in 2000 and 2004." This new venture, I presume, is just a great continuation and progression of that past practice. She was careful to emphasize, as Smuckler had done, that "We will not call the election." To worries that all of this interferes and meddles with voter behavior, she adamantly said "voters are smart, they can handle it."

Turner also sought to limit the visibility the information would attract by saying that "our readers are sophisticated." When I brought up the fact that not just Slate readers would be able to acquire the data, she suddenly had to get off the phone.

There were congressional hearings after the 1980 election. The networks voluntarily made a pledge not to project the final results until all the states had voted (they will still call individual states after those states' polls close, however). They were not legally obligated to do this, but I'm very glad they made that decision. It was the right one. It demonstrated that they were being responsible and did not want to affect the vote in any way.

VoteCastr and Slate, however, are not being responsible. By their actions, they could alter the outcome of the 2016 election. Their motives are to be players and make money and enhance, above all, their respective brands.

Now, I know human behavior is random and unpredictable. I am fully aware that nearly one-third of the U.S. population will participate in early voting (which is available in 37 states and the District of Columbia) and so they are being influenced by the polling results that are disseminated every day up to Election Day. But call me old-fashioned, call me a purist; I believe Election Day is sacred. Anything that in any way alters, disrupts or interferes with it should not be approved nor sanctioned.

Yes, the campaigns will do what they always do on Election Day; that's their legitimate role, and I expect no less from them. VoteCastr and Slate are not doing what they are doing in the public interest. Make no mistake — they are doing this in their interest.

I hope the networks will never change their current policy. This restraint and discipline is good and right for the country and does not tamper with democracy. They should be applauded and supported in every way.

And don't kid yourself: VoteCastr and Slate are attempting to call this election, without actually calling it.

Plotkin is a political analyst, a contributor to the BBC on American politics and a columnist for The Georgetowner.


The views expressed by contributors are their own and not the views of The Hill.