February 22, 2016
We’re going to start seeing a steady stream of March 1 polls over the next couple/few days. Many pollsters waited for South Carolina/Nevada results on Saturday before beginning to sample. One exception is Emerson College, which took a look at Massachusetts on both Democrat and Republican sides from Friday to Sunday.
This lets us engage in a favorite pastime of mine. How many conclusions can we draw from a single poll in a single state from a single pollster? Normal sample size caveats apply, but we can still jump to plenty of reasonable conclusions.Trumplandia
If Emerson is correct, they may wind up renaming the Commonwealth after The Donald. He gets a full 50% share of total GOP support, with Marco Rubio an absurdly distant second at 16%. John Kasich is making the Bay State a priority, but even after spending Saturday there instead of South Carolina, sits at 13%.
You wouldn’t expect Ted Cruz (10%) or Ben Carson (2%) to feel at home there and the poll agrees. Is Trump’s number a ridiculous outlier? No.
Back in October, Emerson had him at 48 percent. There is a dissenting opinion. A Boston Globe/Suffolk poll from November had Trump at a measly 32%, though still leading easily. That’s the sum total of our data sample.
Unlike CNN/ORC, which consistently has Trump at the top of his range, Emerson exhibited no pro-Trump lean when they polled previous states. They were 3 points high in Iowa, 4 points low in New Hampshire, 3.5 high in South Carolina.
This is for real. Not only is Trump likely to win Massachusetts, it’s a place for him to smash through the 40 percent barrier. Plenty can happen in a week, but leading your closest competitor by a 3 to 1 margin is a great place to start.
What about the race for second? It’s possible the gap between Rubio and Kasich is far larger than it appears. I mentioned they polled Friday, Saturday, Sunday. Emerson was kind enough to break this out.
Friday: Rubio 15%, Kasich 12%
Saturday: Kasich 17%, Rubio 13%
Sunday: Rubio 23%, Kasich 8%
I’ll readily concede correlation isn’t causation, and the sample size is miniscule, but it sure seems like they started relatively evenly, Kasich got a small jolt of momentum for focusing on the state, then Rubio got a bigger jump for doing well in South Carolina.
Finishing third would look very bad for either, so in a world where Trump does not win virtually every state on March 1, this is worth monitoring. If there is something to this trend, Kasich got dinged for winding up in the distant third tier, and Rubio is showing his customary swing in support based on whatever just happened.
This isn’t New Hampshire
Emerson had Bernie and Hillary tied at 46. A few days prior, PPP showed Bernie up 7 (49/42). As a New England state with a predominantly Caucasian population, you might expect Sanders to have a bit more of an advantage.
Massachusetts is the one state that opted for George McGovern in 1972. That was more than a couple minutes ago, but it is also home to Senator Elizabeth Warren. The Warren factor might explain a few breakdowns that are different from elsewhere.
Normally the biggest factor in a Clinton-Sanders poll is age, followed by whether the voter is African American. Next is Democrat or Independent. Gender is fourth. In this case, most of the voters are white, and we don’t have a separate breakdown by ethnicity.
Though the Democratic primary is open to Independents, and Sanders has his customary 2 to 1 edge with them, Emerson is projecting a smaller share of the electorate than in New Hampshire. This benefits Hillary tremendously and is part of the reason this is a closer contest.
The percentage of black voters is a bit higher (if still lower than Nevada, never mind South Carolina), which helps Hillary too. Even without the data, I’m going to assume she has the vast majority of African American support. There’s no indication anywhere else this isn’t true.
The age gap is way smaller than elsewhere. Normally, Bernie leads 7 or 8 to 1 with the youngest voting groups. Here it’s 2 to 1. Normally Hillary pulls ahead once voters are north of 45 or 50 and dominates over 65. In this case her large edge doesn’t materialize until voters are 75 or over.
The gender gap is larger than normal. Each have a 25 point lead among their own kind. My theory is Warren has something to do with the age and gender differences compared other states. The following is mostly unprovable, but innocent speculation is relatively safe.
The normal split is Hillary wrapping herself in the Obama Administration blanket, while Bernie plays the insurgent avenger, willing to risk apostasy by criticizing the Democrats’ leader. While I’m sure the president is plenty popular among Bay State Dems, so is Warren.
Older voters have seen Warren grow in stature over the past few years with a platform like Bernie’s. He wouldn’t seem as implausible to them as a similarly aged voter in Nevada. You can imagine older male voters seeing him as a less feminist Warren.
Similarly, younger female voters might be more open to Hillary because of Warren. the senator combines a more directly feminist agenda with Bernie’s economic view. Where Hillary is a bridge too far for Millennials in Nevada, the young women of Massachusetts are potentially more open to her.
If there’s anything to Emerson’s findings, it’s a reminder you can’t just take what you see in one place and apply it evenly across the country. Many states have distinct characteristics.
It leaves us with Bernie lacking the edge he might have preferred in a state viewed by many as must win for him. It was never going to be easy. Hillary defeated Obama handily in 2008, even after he received the Kennedy family endorsement. She ran far ahead of her New Hampshire result there.
If you assign Bernie the normal challenger’s share of undecided voters, he pulls ahead, and led in the PPP poll, so for now, he’s a favorite, but he shouldn’t get comfy. A poor result in South Carolina (relative to expectations) could depress his turnout or push a few more voters to Hillary.