Emerson College poll has Markey up 12 (45-33)

Who isn't interested in polls?? Everybody loves a good poll. :-) - promoted by david

With apologies to those who aren’t interested in polls. Read no further…

The Emerson College Polling Society has a new survey, fielded this Monday to Wednesday (5/20 to 5/22), showing Ed Markey leading Gabriel Gomez in the U.S. Senate race, 44.8 percent to 33.1 percent, with 22.1 percent still undecided. That’s an improvement from their poll taken right after the primaries, which had Markey up by 6. I don’t believe I saw the first Emerson poll, but it fits well with PPP’s poll showing Markey up 4 just after the primary, and the subsequent polling showing him ahead by 7 or 8. Suffolk’s poll showing Markey with a 17-point lead (52-35) remains the outlier, though this new poll shows movement in that direction.

Read on for nuts and bolts…

The Emerson survey used two screening questions. First, “How likely are you to vote in the June 25 Special Election…?” Second, they asked if participants had voted in either primary. I don’t know, however, how they used the responses to construct their likely voter screen.

81% said they were “very likely” to vote on June 25 and another 8% “somewhat likely.” I’d love to see the day we get 89% turnout for a special election in June. 68.6% said they had voted in a primary (40% in the Democratic primary and 28.6% in the Republican primary). If true, that’s much higher than the turnout rate on April 30.

A full 98% now have heard of Markey (early in the primary season polls were showing a third unfamiliar with him). Markey is at 49.5% favorable, 35.6% unfavorable, 12.9% undecided. Only 6% said they’d never heard of Gomez. Gomez is at 41.1% favorable, 34.8% unfavorable (a big rise in unfavorable), and 18.1% undecided (a big decrease in undecided, suggesting people don’t like what they’ve learned). These results could reflect the more engaged sample, but that’s who will show up on June 25. What’s curious is that this poll seems to have more regular voters, but the share of undecided respondents is twice as large as in PPP’s latest poll.

Markey leads among Democrats 66-15 (Gomez pulled 12% of Dems in PPP’s poll last week). Gomez leads among Republicans 70-12 (PPP had a similar result). Gomez’s lead among unenrolled voters is 43-32, with 26% undecided. That’s lower than his lead in the prior Emerson poll or PPP’s polls, and less than he needs to win.

A third of respondents identified “jobs” as the most important issue facing the country. 19.4% named healthcare, 15.9% named education, 14.0% named the deficit, and only 4% named terrorism, a surprisingly low figure only a month after the Boston Marathon bombing. Perhaps learning more about the isolated nature of the Tsarnaev brothers has prevented a full-scale panic on terror.

Relatively little difference in the issues identified by those who voted in the Democratic primary and those who voted in the Republican primary. Nearly indentical percentages of each group named jobs (many) and terrorism (few). Democrats were slightly more likely to name healthcare. The big difference was more Democratic primary voters focusing on education (19% vs. 10%) (see that, Bob Neer?) and more GOP primary voters focused on the deficit (also 19% vs. 10%).

These disparties were much larger among self-identified Dems and Repubs as opposed to April 30 primary voters. 20.9% of self-identified Democrats said education is the most important issue, but only 6.6% of self-identified Republicans. 28.6% of Republicans named the deficit, compared to only 8.3% of Democrats. Just over a quarter of Gomez’s support (25.1%) comes from people who believe the deficit is the most important issue we face, compared to only 5.7% of Markey voters. Guess Fox News hasn’t started reporting that the deficit is largely under control, probably more than it should be in such a weak economy.

Some quibbles:

  • The poll might oversample self-identified Democrats (44.6%) compared to Republicans (10.5%). Independents were 44.9% of those surveyed. Perhaps hard-right Republicans are not passing the likely voter screen.
  • The self-identified Democrats were 54.5% men, 45.5% women. Seems like not enough Democratic women. Women were a large majority (57.6% vs. 42.4%) of the unenrolled voters surveyed.
  • 42% of those polled were in the 60+ age group, a demographic that has been very friendly to Gomez in all the polling thus far. In this poll they favored Markey 42-34. Odd.
  • Only 3.2% of respondents were age 18-29, and only 8.1% between 30 and 39. Perhaps a landline poll?
  • A strange geographic breakdown. Not in terms of numbers (about which I don’t know enough to have an opinion), but in terms of process. They identify regions by area code. Just last week I was telling my wife that it’s crazy to have Salem in the same area code (978) as Dunstable and Ashburnham, but not as Marblehead. Lexington is 781, same area code as Marblehead or Scituate, while Concord is 978. 508 includes Sturbridge and Worcester, but also Nantucket. I find it odd to call 508 “Central/South Shore” when that’s not even a region, and the entire South Shore north of Plymouth is in 781. But I digress. My point is I wouldn’t use area codes to identify regions in Massachusetts.

Massachusetts area code map - WikiCommons

I’m not going to rely too much on any poll, or even all the polls in this race averaged, but the trends continue to be favorable for Markey as we approach the halfway point of the general election season.


5 Comments . Leave a comment below.
  1. 82% likely to vote? 2/3 voted in the primary?

    While I love these results, it’s hard to believe that this is accurately modeling what the turnout would be if the election were held today, let alone predicting results next month.

    I assume they’re using area codes because they were random dialing and didn’t get residential info from respondents.

    • Here's the thing

      I don’t think it’s accurate modeling of the electorate as a whole by any stretch. It might well be more accurate modeling of who will turn out on June 25.

      I agree with your theory on area codes. My thought is that, if that’s all you’ve got, list the area codes but don’t assign any region to it. Not a big deal though.

      • yes take that back about likely voters

        Shouldn’t try to do 2 things at once and comment mobile. That must be their model of likely electorate, of course.

  2. What is the Emerson College Polling Society?

    I’m not familiar with it. Does it have a track record, or is it just a political club that made some telephone calls?

    • Seems to be

      student run, but the methodology is similar that that used by virtually every other pollster out there. It wasn’t students on the phone, the poll was automated.

      They’ve been around a while and have done a bunch of polls, some of which have been reported in the press over the years. This is a (non-exhaustive, I believe) list of prior polls.

« Blue Mass Group Front Page

Add Your Comments

You must be logged in to post a comment.

Fri 28 Apr 11:50 AM