I have spent a lot of time complaining about how the media interprets early polling in the Massachusetts gubernatorial race as horserace numbers, even though only one candidate has greater than 50% name recognition. But how are the candidates doing among the people that do know their names?
I computed the recognition-adjusted polling average for the Democratic candidates for Governor and wrote about it on the Mass. Numbers blog:
http://massnumbers.blogspot.com/2014/04/adjusting-ma-governor-polls-for-name.html
The upshot is that Martha Coakley still has a commanding lead, with Steve Grossman, Juliette Kayyem, and Don Berwick in a group behind, and Joe Avellone trailing.
That being said, we are getting to a point where low name recognition needs to go from being an excuse, to something that needs to be overcome fairly quickly, possibly through a strong showing at the convention.
I thin your site is amazing, and haven’t seen it before until now.
The work you have done on your site is an invaluable resource. I really like the partisan rankings.
When the new Fivethiryeight launched they did away with all the interactive maps. You are doing a better job at sharing the data you have than Nate Silver currently is! The icing on the cake is you are covering local politics, which I actually care about a bit more than the national stuff!
And consistent with what I am hearing when I am out talking to voters. I am a supporter of Martha Coakley and we’ve been out talking to voters door-to-door.
It was nice to see you on Sunday. Thanks for your work on “Massachusetts by the Numbers.”
I’m not really sure what to do with this information. Is the implication that the dynamic of the race would shift if people knew the less-well-known candidates better?
In this case it appears that even controlling for that Coakley has a comfortable lead, but the premise is that it’s not really a fair fight if some candidates are better known than others.
Some key takeaways:
1. Martha Coakley is in the lead, whether you look at the raw numbers, or within the subsets of respondents that actually know the more obscure candidates.
2. Grossman, Kayyem, and Berwick are still in the running, but need to quickly make up name recognition ground—a strong showing at the convention could really help with this.
3. It is a bigger stretch to see Avellone making up enough ground to win.
Avellone and Kayyem fail to get 15%, Kayyem backs Coakley, Avellone backs Grossman, and it comes down to Berwick, Grossman, and Coakley.
If Berwick can get air time, momentum, debates, I think he can close the gap. He may also benefit by Grossman and Coakley going real negative against each other, like Reich benefited when Tolman, Birmingham and O’Brien did in 2002.
But the late primary will hurt us if it goes really negative and saps a lot of energy and momentum. Particularly since Charlie Baker 2.0 is IMO a significantly better campaigner than he was in 10′.
…he still lost to O’Brien by 8 points (an only captured 2nd with 6,000 more votes than Birmingham). Berwick is the Reich of 2014, and I mean that he wins the hearts and minds of the party’s soul, but loses the primary.
Also, while I understand that the primary is late, but in terms of September, it is early, just one week after Labor Day. Tough to get undecided’s engaged one week after the “end of summer”.
IL held its primary in March and we are stuck with another nasty few months of this race out here. That said, it seems like the nominee will win a plurality and ensure a short time table. I’m no Coakley defender, but the short timetable from primary to general in the special didn’t help her.
And my head tells me you’re right on Berwick, but he has my heart and support. Reich really did have students running his campaign ( which shows is how potent his message truly was that he did as well as he did!) while Berwick has one of our most talented campaigners running his (Chang-Diaz) so we will have to see. I won’t defect to Grossman just yet.
How could we change the primary to IRV?
I think there’s no question, in response to JimC’s question, that “the dynamic of the race would shift if people knew the less-well-known candidates better.” I also think there are limits to how much we can glean from this type of analysis.
On the one hand, I’d expect candidates with low name recognition to have a higher proportion of strong supporters among those who’ve heard of them, as compared to a candidate like Martha Coakley. If practically the whole electorate’s heard of you, you’re more likely to have people who don’t like you. That would suggest Coakley’s position may be understated.
On the other hand – and I think this is big – is what I call the “Back Bay” effect. Some years ago I had a friend who moved to Boston from New York. When making the move, he planned to live in the Back Bay. He’d heard of it, he’d even been there, it’s pretty. I thought JP would be a much better fit for him. He’d vaguely heard of JP but was skeptical – the Back Bay was just so much more prominent in his mind and, if I’d asked him just after I first suggested JP, he’d have gone with Back Bay. As he came to learn more about both places his skepticism faded. Now he’s a happy 20-year resident of JP.
The point is this: even some people who’ve heard of Coakley and, say, Grossman, may have heard a lot MORE about Coakley at this point. If they’re asked whom they’d vote for if voting today, they’ll go with the one they know best. That may change when the race shapes up, post-convention.
It’s clear that, ultimately, the non-Coakley candidates will have to become better-known and impress people if they’re going to win. I agree with the idea that a strong convention showing would help, but I think there’s still time after that. Deval Patrick (who, of course, had a strong convention showing) trailed Reilly by more than 20 points in polls into May, had a four-point lead in late August, and beat him by 25 points in September. And nobody needs to win by 25 points. A 2-point win is just as good.
2002 makes more sense since it is also a five person race like that one was, that became a four person race as this one likely will (with Avellone going like Grossman did in that cycle). We also had an established female front runner with a big lead and name recognition vs. another statewide office holder (Senate Prez sort of counts) and two outsiders of sorts (though Tolman is probably better known than Kayyem or Berwick were at this point).
Then it becomes how to snatch the crown, I am strongly hoping the convention can weed out underperformers like Avellone and Kayyem, that is probably the best chance Berwick has to take on the Patrick role since in a three person race with the two frontrunners running to the middle and beating each other up the positive progressive can surge with grassroots support.
Otherwise I see a plurality for Coakley, a bruising primary with Grossman that damages her as the nominee, and Baker getting a 4-5 pt lead. Hopefully 2002 doesn’t repeat itself.
There are parallels to 2002, to be sure, but this year presents a very different climate. Democrats generally were on their heels that year, in the first major post-9/11 election season. This year, it’s the Republican brand that’s crap in Massachusetts. Massachusetts Democrats are far better organized at grassroots election winning now.
I’m not scared of Baker yet, but I continue to think there’s not a chance in hell Berwick’s the next governor. In many ways I wish he would be, but I don’t see it at all.
I think there is merit in your view on name recognition and viability, which reminds me of this paper I found when looking at the academic research around name recognition:
NAME RECOGNITION AND CANDIDATE SUPPORT
Cindy D. Kam and Elizabeth J. Zechmeister
Vanderbilt University
http://www.princeton.edu/csdp/events/Kam04282011/Kam04282011.pdf
ABSTRACT
Does familiarity with candidates breed contempt, indifference, or admiration? Existing work presents modest theory and mixed evidence regarding whether name recognition affects voters’ support for candidates. We develop and test a three-tiered argument, postulating that a) name recognition can affect candidate support; b) a key mechanism behind this effect is inferences about candidate viability, and not inferences about traits or experiences; and, c) name recognition effects are most likely to obtain in low information contexts. Using a series of three laboratory experiments, we show that name recognition can affect inferences about candidate viability, which in turn affect candidate support. The effects of name recognition, however, are vulnerable to the presence of a more applicable cue, specifically incumbency status. The article thus speaks to debates on the effects of name recognition and incumbency as cues and, as well, extends current understandings of the extent to which subconscious influences shape political decision-making.
I will check it out.
Here we don’t have an incumbent governor running, but I’ve long thought AG offered more opportunity to get one’s name before the public than other statewide offices. Martha Coakley also has been, fairly recently, the nominee in a statewide election that had the media field to itself.
that 39% of voters (as of March 21) have never heard of Grossman.