Effect of polls on races

At the end of December, the Boston Herald ran a story headlined, “GOP Lets Scott Brown Fend for Himself.” Just 20 days ago the Herald reported, “Senate candidate Scott Brown has been all but abandoned by the same national Republican committees that pumped hundreds of thousands in campaign cash to … Mitt Romney and William Weld during their long-shot bids for U.S. Senate.”

{mosads}On Jan. 1, Scott Brown reported just $367,000 in the bank — not enough to wage a serious campaign. Between Dec. 29 and Jan. 4 he ran less than 200 points of advertising — a volume guaranteed to be barely noticed, despite a clever ad.

All of a sudden, on Jan. 11, Brown raised $1.3 million in a single day. And each day last week Brown raised over a million dollars — more each day than he had raised during the entire four-month period since Sen. Edward Kennedy’s (D-Mass.) death. Outside groups began adding to Brown’s total.

The press began devoting vastly more attention to Brown’s campaign — and to Coakley’s alleged failings.

What changed? Was Coakley indicted? Did Brown jump into a raging river to rescue a drowning child? No.

What did happen? A poll released on Jan. 9 put Brown in the lead. The very day of Brown’s Internet money bomb — which began with a $500,000 goal and ended with a $1.3 million haul — another poll added fuel to the fire by suggesting Brown was within two.

These two polls, more than anything else, galvanized Republican donors, motivated volunteers to join the Brown effort and thrust him in the media spotlight — helping create the possibility of a Brown victory.

If he had added to his $367,000 at a normal pace, Brown would have been unable to deliver his message; if he had not basked in positive press coverage while leaving Coakley to wallow in a flood of “what’s wrong with her” stories, Brown may well never have moved into the lead.

Make no mistake, the polls alone were not sufficient to produce a Brown wave. If some errant poll in 2006 had shown Kenneth Chase just two points behind Ted Kennedy, would Kennedy have lost? Certainly not. Other factors conspired to make this a winnable seat for Republicans this year.

However, without the close polls, the circumstances that made Republican victory possible would have been insufficient to bring it about. The polls were the spark that ignited the dry kindling on the forest floor. Without the spark provided by the polls, though, there would have been no conflagration.

Is there anything wrong with polls influencing elections? If the polls were accurate reflections of reality, it’s hard to complain. Though we will never know for sure, my own strong sense is that these two IVR auto-dial polls significantly overstated Brown’s support when they were completed.

A Boston Globe poll just before PPP’s gave Coakley a 17-point lead. Our own poll between PPP and Rasmussen showed a slightly lower 14-point advantage for the Democrat (though in our data the race moved continuously and dramatically toward Brown after that).

Moreover, examine the story the polls tell. PPP would have you believe that hardly anything happened in this race during the last week — Brown’s margin went from one to five.

Our polling and the Globe’s suggest Coakley was ahead at the beginning of January by the same margin she had in GOP polling toward the end of December and then, in the last week, the race changed dramatically.

That story seems a lot more plausible than the tale told by the IVR polls.

So here, a false prophesy may have become self-fulfilling.

Mellman is president of The Mellman Group and has worked for Democratic candidates and causes since 1982. Current clients include the majority leaders of both the House and Senate.