Discover more from The Intersection
How independent voters could upend the New Hampshire's Republican primary
It’s been a busy week for book promotion, so no Friday roundup this week. I’d like to take the opportunity instead to write about a brief note about something odd in the media’s polling of the New Hampshire primary.
In book news, check out my appearance on the FiveThirtyEight Politics podcast, the review in The Economist (“‘Party of the People'‘ a book by Patrick Ruffini, a Republican pollster and data maven, brilliantly dissects the changes within the disparate voting blocs in America, combining rich data analysis with vignettes from American history”), video of the book event at the AEI, and a conversation with Puck’s Tara Palmeri.
In 2012, the last year without a competitive Democratic primary, less than a majority of those who voted in the New Hampshire Republican primary identified as Republicans. 49 percent were Republicans, 47 percent independents, and 4 percent were self-identified Democrats.
In 2016, both party primaries were competitive. 55 percent of those who voted in the GOP primary were Republicans, 42 percent independents, and 3 percent Democrats.
What do polls this year expect the Granite State Republican primary electorate to look like?
Monmouth has it at 83 percent Republican identifiers, 17 percent independents or Democrats.
CNN/UNH has it at 75 percent Republican identifiers, 25 percent independents or Democrats.
This is wildly off from past historical patterns, at least if you believe the exit polls. Current media polling expects way more partisans and way fewer independents than we have seen turn up in New Hampshire historically.
The Intersection is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.
And this could matter quite a bit for how candidates place. While Donald Trump has a strong lead among self-identified Republicans, he takes just 12 percent of independents. Chris Christie takes 36 percent and Nikki Haley takes 31 percent of those votes. Haley performs more strongly in this poll overall because she doubles Christie’s support among Republicans.
Meanwhile, Trump takes just 7 percent of independents and Democrats in the CNN/UNH poll. Christie has 37 percent of these voters, with 28 percent for Haley. Haley is second place in the poll thanks to actually having an identifiable constituency among base Republicans.
So, yes, whether independents are 20 percent of the New Hampshire electorate or something like 50 percent seems to be a question that matters a great deal.
Wonky side-note: CNN also provides a breakdown of the voter file registration of respondents—and this is a more reasonable 57 percent Republican to 43 percent independent/unaffiliated. This is close to what I’m expecting the real breakdown to be. But their percentage of self-identified Republicans is much lower compared to previous exit polls.
This disconnect suggests a clear path for a moderate-lane candidate to outright defeat—yes, I said defeat—Donald Trump in New Hampshire.
Right now, Chris Christie is in the role of the spoiler. He has little crossover appeal to the Republican base—what an odd sentence to write about a Republican candidate!—and based on his horrendous favorability ratings among Republicans, no ability to sustain momentum into South Carolina and Super Tuesday. His continued presence in the race is a gift to Donald Trump.
New Hampshire has given us some wild results before—and it could again
I was in New Hampshire for the 2000 Republican primary for George W. Bush. The final polls had us trailing by a point or two to John McCain. Bush lost by 18 points, thanks to a surge in support for McCain among independents.
The furthest-left candidate in the field tends to get traction because of the substantial crossover vote, including a large number of votes from Democratic-leaning independents. (This will be exacerbated by Democrats not having a real primary, with zero delegates at stake and Joe Biden not on the ballot.) So, John Kasich emerges in second place in 2016 compared to more nominatable alternatives. Jon Huntsman gets some traction in 2012. And the 2016 Trump has more appeal to more apolitical independents, allowing him to run the table that year.
When races are competitive, turnout in New Hampshire presidential primaries is exceptionally high, above midterm-year levels. That includes participation from even low-information independents. And Trump is exceptionally weak among these voters.
Of course, the conditions a sneak attack on Trump that exist in New Hampshire exist almost nowhere else. Even in states with open or semi-open primaries, turnout isn’t that high, so Republican voters are functionally in control of those contests. As such, anyone who emerges from the first two contests would need to be able to grow support among Republican base voters now firmly in Trump’s camp.
Media polls are in a tough spot in this, and I have every confidence that they’re doing their best. They have to poll both primary electorates, with a roughly 50:50 divide between those who plan to participate in the Republican or the Democratic primary. But the strong likelihood is that there won’t be much drama in the Democratic primary—unless Dean Phillips somehow emerges as a serious threat and there’s a concerted push to get write-in votes for Biden, a possibility not to be entirely discounted. But assuming no real competition on the Democratic side, votes in the respective party primaries skew heavily towards the side with the more competitive contest. It was 4-to-1 Republican in 2012 and 2-to-1 Democratic in 2020.
As a result, media polls are not correctly capturing the actual decision independent voters will make on primary day as to whether to pull a Republican or Democratic ballot. And that could lead to some surprises.
Update: Monmouth’s Patrick Murray clarifies that their survey uses leaned party identification while the exit polls use unleaned party ID, so the two numbers are not directly comparable. CNN/UNH, with a higher percentage of pure independents and Democrats, shows about a 3:2 ratio of respondents likely to vote in the Republican and Democratic primaries, which is below the historical norm for parties with competitive primaries, so we can expect an electorate that skews further in the Republican direction on primary day. Monmouth didn’t seem to have polled the Democratic primary, so I do not know what this ratio would look like for them.