Deval Patrick's lead may be bigger than you think

Friday, the Boston Herald screamed the headline "Poll says race is between Cahill & Baker." It's safe to say that a Herald reader that didn't get into the article would believe that Governor Deval Patrick was getting trounced, or at least was losing. Funny, that's not what the poll said:

Gov. Deval Patrick’s standing with voters is so weak that this year’s race for governor is shaping up as a contest between his two rivals, a new Suffolk University-7News poll shows.

“This race is really between Charlie Baker and Tim Cahill,” said David Paleologos, director of the Suffolk University Political Research Center, which conducted the poll. “Whoever emerges between the Baker-Cahill race is likely to be the winner.”

As for Patrick, Paleologos said, “On paper, he leads.”

The governor’s grip on a precarious lead continues, with the incumbent Democrat taking 33 percent of the vote compared to Republican Baker’s 25 percent and Treasurer Timothy Cahill, running as an independent, close behind with 23 percent.

Green-Rainbow Party candidate Jill Stein trails with a distant 3 percent, and 16 percent are undecided, according to the survey of 500 likely voters.

Amazing. Governor Patrick holds a lead outside the margin of error, yet the "race is really between Charlie Baker and Tim Cahill." In the official Suffolk University press release touting the poll there is no suggestion that Patrick is losing, but there is this note on momentum:

Patrick (33 percent) still leads the tightening field, followed by Baker (25 percent), who edges out Independent candidate and State Treasurer Tim Cahill (23 percent). Green Party candidate Jill Stein has 3 percent, while 16 percent are undecided.  In a November, 2009 poll, Patrick led Cahill 36 percent to 26 percent, while Baker, the former Harvard Pilgrim chief executive, was a distant third with only 15 percent.

Now, it looks to me like Patrick's and Cahill's standings haven't changed a bit (their three-point drops would be within the margin of error), but that Baker has picked up 10 points from undecided voters. I'm still not sure how that translates into a Patrick loss, but there you go.

I suppose seeing the incumbent governor at 33% could mean that he is toast. That is one way to look at it. But there are also a couple of other ways of looking at the Suffolk poll.

For instance, looking at the demographics of this poll, one thing stood out to me. According to the marginals posted on Suffolk's web site, the polling sample went 49-36 for Scott Brown in January's special election. That corresponds to a 57-42 advantage when you take out the 14% who did not vote or refused to answer. We know that Scott Brown actually won by a 52-47 advantage, so I wondered what would happen if I weighted the votes for governor based on a 52-47 margin among the 431 respondents who said they voted. (I did not re-weigh the votes of the other 69 respondents).

Here is how the poll looks based on a 52-47 Brown win:

Patrick -- 35
Baker -- 24
Cahill -- 22
Stein -- 3
Undecided -- 16

That changes the Patrick margin from +8 to +11. Not a huge change, but one that could cause a tabloid like the Herald to scream "Patrick cruising with double-digit lead" if it were inclined to ever say anything positive about the governor.

Another way one might choose to look at the poll is that while Patrick continues to do very well with Coakley voters, neither Cahill or Baker have broken through with Brown voters. Here are the numbers (no weighting):

Among Coakley voters:
Patrick -- 62
Cahill -- 15
Baker -- 7
Stein -- 4
Undecided -- 12

Among Brown voters:
Baker -- 42
Cahill -- 28
Patrick -- 11
Stein -- 2
Undecided -- 16

If Baker is going to match Patrick's base, he'll need somewhere around 57% of Brown voters to come even. He also cannot afford to give up Brown voters to Patrick (11%) at a higher rate than Patrick bleeds Coakley voters to Baker (7%).

Here is another way to look at it. Let's take Suffolk's numbers in this poll and compare them to the six previous polls in this race (there may be others, but I could only find seven):

Notice Suffolk's numbers. In their previous two polls, Patrick had his two best showings while Baker had his only two results under 20 percent. In other words, Baker may not have momentum at all; rather maybe the Suffolk poll has been out of sync in its previous iterations and is just now falling into line with the other surveys.

Sure, it would be great if Patrick were polling in the mid-40s, but he's not. Even so, no matter how you slice it, he is winning today just as he has been since the start of the campaign. If Cahill stays in the race and Patrick shores up the rest of Coakley's supporters, he will be reelected.