Pre-election survey data that the Romney for President campaign shared with a reporter from the New Republic helps explain the Republicans’ excessive optimism prior to the Nov. 6 vote count — but not entirely.
Overly rosy polling data from key states like Colorado, New Hampshire and Iowa make it more understandable why Mitt Romney and his aides thought they had a good chance of upsetting President Obama.
Yet, as the New Republic’s reporting shows, the Romney camp still had to convince itself that late "momentum” (which proved mostly non-existent) would push the Republican to a key victory in Ohio. Obama won the Buckeye State 50.6% to 47.6%, rendering the rest of Romney’s coulda-wouldas somewhat moot.
Neil Newhouse, Romney’s chief pollster, told the magazine (its online reports appear at TNR.org): “I’m not sure what the answer is,” as to why the campaign’s internal surveys misjudged several key states. The misfires were most pronounced in Colorado (where internals had Romney up 2.5% on the final weekend of campaigning and he lost by more than 5%); New Hampshire (where a perceived 3.5% lead dissolved into a 5.5% loss); and Iowa (where Romneyites thought they were in a dead heat and lost by nearly 6%).
The positive convictions were so strong that Romney wrote no concession speech. One staffer described seeing the candidate’s oldest son, Tagg, on election night in a “complete state of shock.”
Those smaller states could not have created a GOP victory but the campaign had also convinced itself that it had locked up North Carolina, Florida and Virginia. Romney ended up winning only North Carolina. (Newhouse did not share his pre-vote surveys for those states.)
The TNR.org analysis points out several other flaws in the Romney camp’s thinking. They assumed, for instance, that undecided voters would break heavily for the former Massachusetts governor. Instead, about 4 in 5 last deciders went to Obama.
As has been amply reported before, the Republican’s model envisioned an older and whiter electorate than turned up at the polls on election day. An expanded Latino voting force explained some, if not all, of Team Romney’s miscalculation.
Finally, the former governor’s aides had felt all along that the most intense voters — and therefore those most likely to vote — were on their side. While many of them were, there were also a number of other voters — younger and from minority groups mostly — who had not registered high enthusiasm earlier, but who came out when it counted.
Pollster Newhouse seemed still at a loss, TNR.org reported, as to why the surveys he used showed late Romney momentum over the pre-election weekend, which did not materialize. He suggested that the miscalculations may have been due to “Sunday polling” — a common concern among pollsters about getting skewed results when sampling over weekends.
As much as they invoke high science, polling models ultimately involve some guesswork when it comes to determining who will show up to vote. Team Obama guessed that it would be able to assure its voters were heard -- in large measure because of the carefully constructed apparatus to get them to the polls. Romney's aides guessed that broader trends would carry them. They guessed wrong.