Suffolk University is out with a New Hampshire poll conducted November 16 to 20 showing Mitt Romney leading Newt Gingrich for the Republican nomination by a 41%-14% margin. Ron Paul also had 14%. That’s statistically indistinguishable from the 40%-11% Romney over Gingrich margin in a Bloomberg News poll conducted November 10 to 12 earlier in which Paul had 17%. And Romney’s percentage is very similar to the 38% to 40% he polled in six of seven New Hampshire polls conducted in October; in the seventh he had 40%. Gingrich was in single digits in those seven polls.
The Suffolk and Bloomberg results are flatly inconsistent with the Magellan poll conducted November 15 and 16 showing Romney leading Gingrich by a statistically insignificant 29%-27% margin. It looks like the Magellan poll was simply wrong, which sometimes happens; polling theory tells us that one out of 20 polls will produce results that are farther away than the margin of error from what you would get by surveying the entire relevant population. So for the moment Romney’s New Hampshire firewall seems to be in place. But if I were the Romney camp, I’d still be a little nervous about this.
