Monthly Archives: January 2010

So much so that I’m having trouble believing them.   We now have another full day’s worth of polls, most of which were in the field at least partly during Obama’s visit, and – keeping in mind all the caveats I mentioned about polling a special election – they indicate that Scott Brown is poised to win this race.  If so, it will be the biggest political upset of my lifetime.

First, let’s summarize today’s polls, including two that came since my last posting.  The first is by Insider Advantage, and it has Brown up by 9%, 52%-43%, with Kennedy taking about 2% and no opinion polling 3%.   The second is by ARG, and it has Brown up by 7%, 52%-45%. Four days ago ARG only had Brown up by 3%, 48%-45%, so Brown has picked up 4% in the last three days by this poll.  Five of the last six polls now have Brown above 50%, with the only exception coming from the Daily Kos poll.

In parsing the details of these polls, there are some rather extraordinary results. Let me start with the InsiderAdvantage poll.  In it, Brown is winning in all age groups except those over 65, with his biggest margin coming in the youngest, 18-29 year old cohort. He’s winning almost 25% of Democrats, and he’s beating Coakley among women (53% of the sample) by 49.5%-46.1% and is clobbering her among independents (37% of the sample) 68.7%-28.5%.  Indeed, some of the results I find difficult to swallow, although in Insider’s defense, results for smaller subsamples can often be all over the map.  Nonetheless, Insider would have one believe that Brown is winning 27% of the black vote and 77% of the Hispanic vote. I suppose it’s possible.

The ARG telephone poll of likely voters was in the field from Friday through Sunday and also has Brown dominating among independents 64%-32%, and winning 23% of Democrats.  However, it shows Coakley with a slim lead among women, 50%-46%. Once again, Brown gets strong support with younger voters, leading 53%-43% among those under 50 and by a smaller 51%-46% advantage among those over 49 years old.

Both the InsiderAdvantage and ARG polls were, at least in part, in the field yesterday, and presumably should have picked up any surge toward Coakley inspired by Obama’s visit.  All told there have been five polls that were at least partially in the field yesterday, and two that exclusively polled on the day of Obama’s visit. None show evidence of any surge toward Coakley – indeed, reaction has been in the opposite direction, toward Brown.   I am not going to say that this reflects opposition to Obama, since these polls include days before Obama’s visit and even for Sunday’s  polls the timing is in question.  But clearly his visit hasn’t – yet – had the impact for which Coakley hoped.

So what are we left with?  One can still construct an argument for Coakley winning this race, and indeed you can find that argument on other websites.  It usually is based on a combination of extraordinary turnout by Democrats tomorrow combined with a response bias in the polling that leads them to collectively underestimate support for Coakley.   For what it’s worth, I’ve made the argument that the response bias may work in the other direction, in Brown’s favor.   Note that the InsideAdvantage poll was still another one using automated polling that gave Brown a larger than expected lead.

Let us assume for the moment that the survey data is correct, and that Brown is poised to pull off the biggest political upset in recent memory.  The question becomes: why?

I’ve been giving this a lot of thought and – not surprisingly to long time readers – I don’t think the standard explanation posited in most major news outlets that this is a reaction to some combination of opposition to health care and to Obama is correct.  Instead, I want in my next post to offer a different explanation for why Brown is winning.  But before I do, I want to give you a chance to chime in.  I know this election has attracted extraordinary interest, because I’m getting more hits today than on any previous topic, save for Election Night 2008.  But most of you are giving me your views through private emails, rather than through the comments section.  This leaves everyone else out of the discussion.  So go ahead – tell me what you are hearing, or seeing, or reading that might explain why Scott Brown is on the verge of pulling the greatest upset since the Red Sox came from three down to beat the Evil Empire.

I’m working on a longer post that tries to get a handle on what we might expect for turnout tomorrow and who that might help, but in the meantime one more poll came in so I thought I’d send it along.

At first glance, it looks like good news for Martha Coakley. The poll, which was sponsored by the Daily Kos, has her in a dead heat with Brown at 48%-48%.   That’s the good news.  The bad news is that the same pollster had Coakley up 49%-41% one week ago, so it is actually consistent with the trend toward Brown noted in my previous blog.  Moreover, this uses live interviewers which may – may – mask voter sentiment for Brown. Ideally, given the house effects of this particular polling outfit, Coakley supporters really wanted to see movement in the other direction.

To anticipate my next post, both Orion and Alex wonder whether Coakley is going to benefit by the apparent nationalizing of this race, since it will send a wakeup call to Democrats that this long-time Kennedy seat is in danger of going red.  One week ago I would have agreed with them that lower turnout favors Brown.  Now, as I begin to gauge the depth of anger among independents, I’m not so sure.  I hope to develop this point later this afternoon.

Meanwhile, news narratives are suggesting dueling frames for the Obama visit, with some accounts focusing on the strong turnout for his speech, and others contrasting it with the turnout for Brown’s counter rally headlined by – you guessed it – Curt Schilling.   Crowd estimates are notoriously unreliable, so I’m not even going to begin to parse this debate.  But it is a reminder that this late in the day, the media airwaves are saturated with ads, claims and counterclaims, which makes it difficult for any single event or message to dominate.  It really is going to come down to turnout.

More in a bit.

Addendum: ARG just in with their final poll: Brown 52% Coakley 45%.

Is Brown pulling away?  There’s conflicting evidence in that regard.  First, a one-day automated poll conducted by the Merriman River Group (MRG) on Friday shows Brown leading Coakley by 50.8% – 41.2%, the second poll in the last four days to show Brown with almost a double-digit lead. Again, however, it is by a less well known polling firm and the cross tabs are not available for me to review.  They do provide basic demographic evidence regarding the polling sample, and nothing stands out as obviously out of line with other polls, except for a slightly higher proportion of women – 59% – than I have seen in other polls.  Otherwise the partisan breakdown seems pretty consistent with what I’ve been seeing elsewhere.  The big unknown, of course, is what their voter screen kicked out of the sample.

The poll finds that Coakley’s support among women continues to slip; she and Brown now split the women’s vote 46%-46%.  The other interesting finding is that Brown’s supporters are mostly concerned about the economy, particularly jobs and taxes, while Coakley’s are focused on health care.  Consistent with earlier polls, the endorsement of Coakley by Kennedy’s widow hasn’t helped, and may have hurt, her campaign, with 27% saying it made them less likely to vote for Coakley, while only 18% said it made them more likely to vote for her.

This poll has a slightly higher percentage of undecided voters – about 6% – than other recent polls, and they are equally divided on whether health care or the economy is the most important issue.

The poll reinforces a point I made earlier, but which bears repeating:  Brown tends to do better in polling done through automated means as opposed to live interviews.  Some media outlets, such as the Boston Globe, have hinted that automated polls are less accurate. Actually, there’s no evidence that I know of supporting that assertion.  Indeed, an argument can be made – and I’ve made it in the context of this race – that automated polling may actually give a better preview of Tuesday’s outcome if people find it socially desirable to say they are supporting a Democrat to replace Kennedy, when in fact they are secretly planning on voting for Brown.  They are more likely to admit to supporting Brown if it involves pressing a number on a telephone pad than if it requires talking to a live interviewer.  If I get a chance, I’ll run the polling data to see just how much better Brown is doing on automated polls.

But that doesn’t mean this particular poll is accurate.  While it’s not impossible that Brown has netted 5% in the last 48 hours, I would feel more confident about this if PPP, which is releasing their two-day poll tonight, comes up with the same results.   An early tease by PPP’s polling director last night, however, suggests that the difference between the two candidates remains within PPP’s margin of error, indicating that the race remains essentially tied.  We should know in a few hours what PPP has found and that will give us a better perspective on this one-day poll as well.  For now, however, I remain skeptical that Brown has in fact opened up an almost 10% lead, as this poll suggests.

Note that all this polling is taking place before the Schilling gaffe, and even the PPP poll will not capture the effects of the Obama visit nor, I don’t think, the recent Coakley rape flier.

I’ll be on later when the PPP poll comes out. To whet your appetite, here are some other teases from last night’s PPP blog:

“-The electorate is becoming more Democratic. Last weekend we found it at Obama +16 and now we see it at Obama +20. So all the efforts to get the party base more engaged in the election are paying off.

-Balancing that out to some extent though is that we’re now seeing Brown win about 19% of the Obama vote, in comparison to 15% on our poll last week.

-Both candidates have seen a pretty large increase in their negatives over the last week, reflecting the increasingly nasty nature of the campaign.

-Even though the race is too close to call overall, 58% of voters think Brown has made a strong case for why he should be elected while only 41% say the same of Coakley. That speaks to voter perceptions that Brown has run the superior campaign and again you have to wonder how different things might be if Coakley had acted with a sense of urgency ever since the primary.”

There, do I have your curiosity piqued?   Now go call your Mom.