Over the last 24 hours, there have been two additional polls. The results are somewhat mixed: The PJM poll has Brown up by 15 points, but the Blue Mass Group has Coakley up 8. How can this be? RCP has a nice roundup of the Mass race here.
One thing nice about both these polls is that they publish the question asked and the percentage of the response. Let’s look at both polls before making some other observations. The BMG poll has Coakley 49-41%. One thing different with this poll is that BMG/Research 2000 did live interviews. This might cause some swing in the percentages when compared to Rassmussen’s polling because he does robo call polling. Politics Magazine points out some of the advantages and disadvantages of each polling technique. BMG seems to be saying that their live poll is more reliable than the robo-poll. That is debatable. Here is a comparison of the robo v. live debate looking at the recent NJ race. Bottom line is both approaches may produce accurate, and inaccurate, results. In the NJ race, however, the robo call poll better predicted the outcome of the election.
The PJM poll used robo calling to generate their poll that shows Brown up 15%. I am skeptical that this poll is accurate, though it may have captured some of the surge heading Brown’s way. As it applies to both polls, I wonder if these results are the “raw” results. They certainly look like it. One thing a poll needs to do to be accurate, is do some predicting of the turnout of the vote. This is notoriously difficult to do in a special election, and getting the formula wrong makes for a bad prediction. Both polls ask a slightly different question in terms of who the voters will vote for on 19 January. So, I think that does not account fofr the variation. Both polls use the likely voters question to narrow the reliability of the poll (predict the outcome). One real difference is the PJM polls about twice as many as the BGM poll. That does not necessarily make the PJM poll more accurate though.
Coakley has very high negatives. When she criticized Scott Brown for shaking hands of Winter Classic fans in the cold outside Fenway, it caused quite a gaffe storm. Curt Schilling (A Republican) criticized her on his blog.Ths could explain this further gaffe (below) when she said Schilling was a Yankee fan.
It is difficult to gauge if this harms Coakley, but I would say it certainly does not help. Whether she thought it a joke or not does not matter, the joke did not work, and it instantly put her into an argument with Schilling–a distraction no campaign needs.
In an effort to revive her campaign, President Obama has made a robo call ad on her behalf:
This may help Coakley since Obama is still rather popular in Massachusetts. However, things look pretty bad for Coakley’s campaign, especially when supporters begin to write critical reviews. This is a signal that Democrats are abandoning her–which suggests that internal polling challenges the BMG poll: Coakley is losing the election.
What have we learned this week? A Brown victory is not unthinkable anymore. I still think Coakley has the advantage because of the Democrat ID outstrips the Republican ID in the state. BUT, it all depends on who turns out. Tuesday night we shall know which polls were most accurate, and which ones deserve not our trust in the future. This is why we have elections, because of all the polls, voting is the thing that really counts.