The race marches on, the candidates keep campaigning, advertising, and debating, and the numbers keep getting worse for Charlie Baker.
Western New England College has released a new poll of 400 likely voters taken October 16-21. Bottom line: the same lead that Deval Patrick has held for months, of roughly 6-8 points, is stable or increasing. Baker has a very slight edge among independent voters (32-26), which is far less than what he needs to overcome the Democrats’ 3:1 advantage in voter registration. Recall that Scott Brown’s overall win was 52-47, and pre-election polling showed him with a 30-point lead among independents. A modest 5-10 point lead among independents is not going to do it for Baker.
Cahill is at 8%, Stein at 5%, and 8% were undecided (5%) or refused to state a preference (3%).
Another piece of good news for the Patrick camp is that among registered (as opposed to “likely”) voters, Patrick’s advantage is much larger, at 45-32. So if the massive Democratic GOTV effort that is now underway has some effect, the odds of getting some of those less-than-super-reliable voters to the polls increases, and should help Patrick.
To be sure, there are warning signs in this poll for Team Deval. One is that the number of supporters who say they might change their mind is slightly higher for him (26%) than for Baker (18%). Another is that fully 69% of Cahill voters said they may change their mind, and if they move to Baker in large numbers, that could be a problem. At this point, I think it unlikely that that will happen; I think the movement will pretty much split along the same lines as the independent voters generally. That won’t be enough to give Baker a win if the preference numbers in this poll hold up.
But the overall message, I think, is clearly “stay the course.” The campaign is working. Patrick’s numbers are slowly creeping upward (Patrick led 39-33 in WNEC’s last poll a month ago), even as a majority of MA voters still answer “wrong track” to the “right direction/wrong track” question. That suggests that people are concerned about the state’s future, but that they also think that Governor Patrick the better person to lead the way.
Next week’s debate should be fun to watch. Baker is going to have to try to shake something up.
<
p>I’m mostly looking forward to Cahill saying something like, “As the father of four daughters, I wouldn’t be there….and Charlie, you went and shared the stage with Jeff Perry.”
The Democrats do hold about a 3-1 registration advantage, but registered Unenrolled voters outnumber Democrats and Republicans together.
<
p>But most of the Unenrolled lean one way or the other, and in this poll, the “Independents” are split into leaning Dem, leaning Rep and not leaning” with the leaners getting folded into the party ID. The split is 234/155/79 D+/R+/indy. So the registration advantage is really a lot closer.
<
p>All of this notwithstanding, I agree with the rest of your analysis and your conclusions. The bottom line on Baker is that he’s not all that likable. He hasn’t closed the deal with enough independents, and if Cahill were not in the race, his vote might split pretty evenly. Stay the course.
How do you know that the poll rolls unenrolled voters who lean in with party ID? That might explain what seems like an oddly small number of independents (which I assumed was synonymous with “unenrolled”) in the sample, but I don’t see the procedure you describe anywhere in the poll documents.
The link you give gives a link to the poll (pdf):
<
p>Right in there is a breakout by “Party ID (including leaners)” and gives the number of people polled in the group (the “N” column).
<
p>There are a couple of recent polls for comparison. The Suffolk poll last week is here (pdf) – it gives a breakout by party ID only (185/61/254 D/R/U).
<
p>The Globe poll from a couple of weeks ago is here, which also breaks out by party registration (168/64/255 D/R/U) and then by party identification (241/162/101 D/R/U).
The Suffolk poll breaks out by party registration only.
The other good piece of news for Patrick is that he actually gains the most support from those aged 65+ and from college-educated voters, which has also been picked up in other polling (see the recent Suffolk crosstabs, for example).
<
p>Since older and more educated voters are the most likely to actually vote come Election Day, that’s a good sign for Patrick.
See my comment here for a report on a GOTV rally in Pittsfield tonight.
<
p>Out here in the Berkshires, where we love Deval and Tim, the issue is not the % of votes they and other Democrats will get (e.g. I believe my town was the best % for Suzanne Bump in the Primary, at 90%), but HOW MANY voters will show up at the polls.
<
p>That is why we are having 3 large canvasses in the County tomorrow, and plan to step up the GOTV effort over the days remaining. Let’s bring this one home, folks!
Along the lines of the conversation between Steve Stein and myself upthread, I emailed the director of the WNEC Polling Institute, Tim Vercellotti, for more information. Here is his very interesting and helpful response, which explains why the poll’s breakdown of Dem/GOP/Ind voters doesn’t match party registration data.
<
p>
It would be interesting to see similar info for the Suffolk and UNH/Globe polls. Their demographics are slightly different. Just a point or two, but if you’re looking for trends, a point or two is important over the space of a week or two.
Suffolk weights based on party registration, while the Globe poll asks both the registration and “how do you lean” questions. The Globe results are interesting, because they suggest (not surprisingly) that somewhat more independents consider themselves closer to the Republicans rather than the Democrats, though not by a huge margins.
to see party tendencies and trends in voting by independents. Through most of my voting life, dating back to 1970, I was registered as an independent, or unenrolled voter. However, in spite of that fact, my candidate choices and tendencies were pretty faithfully Democratic, leading me to decide to register Democratic. I suspect that for most unenrolled voters, the same thing holds true, with votes going pretty much for one party or the other on a steady basis.
Although I’m a bit older than you, I guess (I’ve been voting faithfully since I turned 21, which was the voting age in 1967).
<
p>I registered as an independent (later that was changed to “unenrolled” when Ross Perot started the Independence Party, or some such name, if memory serves). I figured I wanted to vote for the best candidate, no matter from which Party they came. Those were the days when the GOP and the Dems were not divided on social issues, only on fiscal ones. Think Rockefeller, Lodge, Saltonstall, and the like.
<
p>I voted for John Anderson some 30 years ago.
<
p>The landscape began to change, however. As the GOP moved further to the right, my interest in their candidates dwindled. Ross Perot’s candidacy allowed a conservative Democrat to take the White House. I was never a big fan of Clinton (though I voted for him) because his “triangulation” politics were far to the right of me.
<
p>My moment of fury came, though, when Clinton was indicted. On that day, I vowed I would NEVER vote for another Republican, and I changed my party affiliation. I’ve been a loyal Democrat ever since.
<
p>Perhaps the Bay State has so many “unenrolled”s because people have just not gotten used to the great divide that now exists between the parties. Perhaps some of them want to be spoilers by voting in the opposite Primary.
<
p>I think it’s time we had a closed Primary system in Massachusetts.
It’s always a good thing to see a polling firm be open with their methodology. I’m on board with the way the WNEC Polling Institute does things, because plenty of research has indicated that the number of “true” independents is actually quite small. Some independents votes even more faithfully for one party than some party members do, but choose to remain independent for a variety of reasons.
<
p>Compare this to the Suffolk approach [PDF], which asks the following question:
<
p>
<
p>Suffolk does not ask about which way independents lean. Suffolk’s a good poll, especially when polling here in MA, though their party weights, which reflect the party registration breakdown (37D/12R/50I) underestimates the percentage of Republicans that vote in midterm elections and overestimates the number of independents. The Democratic number is about right (though very slightly lower than it will be), but Republicans will be closer to 20% and independents closer to 40% of the November 2nd electorate.
<
p>The Boston Globe poll [PDF] asks both the registration question and “do you consider yourself more of a Democrat or Republican” question, and present the results (somewhat confusingly) as, for example, “registered Democrats” and “Democrats”, the latter meaning Democrats plus independents who lean Democratic. The Globe data thus provides a good head-to-head comparison between the two.