Sure, it was only by eight votes. And the vote count probably wasn’t accurate. And it has no bearing on the actual number of Iowa delegates he will get. But Mitt Romney won the Iowa caucuses. Just a few weeks ago, this seemed highly improbable.
So what happened? A stealth organizing campaign and oodles of PAC spending on TV ads tearing down his opponents certainly played the major role. But the on-the-ground presence of New Jersey’s governor in the final days before the vote will certainly be seen as a factor – certainly by Chris Christie’s supporters.
It’s difficult to say whether many actual caucus goers were swayed by Christie’s campaign appearances. But he went there. He got good press. And Romney won. Favor owed. That’s good for Christie and good for New Jersey.
In the end, Romney did a couple points better than the 23 percent he was averaging in the final pre-election polls leading up to the vote. This was just two points shy of the 27 percent win I predicted for him two weeks ago.
All the other candidates, but one, performed about a point or two below their pre-election poll numbers. The exception was Rick Santorum, who finished more than eight percentage points above his final pre-election average of 16 percent.
Basically, each of the other candidates lost a little support to Santorum in the final days and he and Romney split the remaining undecided vote. It’s worth noting that Santorum did not hit double digits in any poll until 15 days before the caucus. He was steadily gaining support at the rate of a couple points a day in the final week. The last poll interviews were conducted on Jan. 1. If polling had continued up to the caucus itself, he could have conceivably ended with a 23 percent showing in pre-election polls.
So the polls were not off the mark. The major bone of contention over the pre-election polls was whether Ron Paul’s support was being accurately represented. Many pundits noted that his firmly committed supporters were younger, and thus less likely to be included in a standard landline telephone survey. I argued the opposite, that while they may be committed, the polls had too many independents in their sample and were thus over-representing the number of Paul supporters who would actually show up at a caucus.
In the end, it was neither. The final poll average for Paul was 21.5 percent. He got 21.4 percent of the vote. He won the independent vote with 43 percent. This was lower than some polls had shown him doing among this group. While 23 percent of caucus goers identifying themselves as politically independent was higher than the 13 percent who did so in 2008, it was also lower than some pundits had predicted. In the end, the polling on Paul’s support was pretty much on target.
Interestingly, Iowa GOP voters who said the most important candidate quality was being a “true conservative” split their vote between the libertarian Paul (37 percent) and the socially conservative Santorum (36 percent). This underlines a key issue in polling on political ideology – the meaning of these terms are in the eye of the beholder.