Biden's working class New Hampshire woes
What we can learn from the first Democratic ballots of the 2024 race
Much of the attention to the primary has been justifiably focused on the Republican contest, but there was a Democratic primary in New Hampshire on Tuesday. Joe Biden won it as a write-in candidate with 65% of the vote.
While this result will likely be enough to calm any speculation that Biden faces any kind of serious threat from Minnesota congressman Dean Phillips, demographic patterns in the results confirm Biden’s vulnerabilities with working class voters that we’ve been seeing in the polls for months now. For just as long, the critics of this kind of polling analysis have been asking to look at “actual votes,” specifically the results of abortion referenda and special elections where Biden himself is not on the ballot. Well we now have “actual votes” in the presidential election, and though Biden’s name was on the ballot, interesting patterns have emerged in the pattern of the votes cast for and against him.
The Intersection is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.
Specifically, there was a large divide in the Biden write-in vote by income. In wealthier and more well-educated towns, Biden regularly topped 70% of the vote. In Hanover, home of Dartmouth, Biden got 79%. Meanwhile, in working class areas with lower average incomes, Biden often struggled to break 50%. The relationship is pretty strong, as you’ll see in the chart below.
New Hampshire is not very diverse, and is one of the toughest states to model because different areas tend to be pretty homogenous demographically—nearly all white, with similar levels of college education. But even then, the income divide is too big to ignore. And where you did have an Hispanic presence, particularly in Nashua’s Wards 4 and 6 and Manchester’s Wards 5, 7, and 9, Biden typically commanded less than a majority of the vote. (Many of the voters here are white, though, and this is probably an extension of Biden’s general weakness among blue collar voters.)
In Manchester’s working class Ward 9, for example, Biden got just 41% of the vote. In Nashua’s Ward 4, the most Hispanic in the state, Biden got 49% of the vote compared to 65% in neighboring Ward 1, relatively speaking the whitest and best educated in Nashua. Turnout was relatively low in working class areas, but that too is a warning sign in areas that still vote Democratic in general elections (and Ward 4 does).
But that wasn’t all. Opposing party candidates often do get write-in votes in New Hampshire primaries, so there were votes for Donald Trump, Nikki Haley, and others in the Democratic primary. In an unsanctioned primary, where most of the action is on the Republican side, votes from registered Democrats motivated to show up in the primary for Republican candidates are a particularly strong signal of likely defection in the general election—even if this sort of cross-party protest if a peculiarly New Hampshire tradition.
And the relationship between Trump’s write-in support and Biden’s runs in the opposite direction. Statewide, Trump got 2% of the Democratic write-ins (less than Haley.) But in Nashua’s working class Ward 4, he got 5% of the write-ins. (Amongst the students and academics in Hanover, Trump got zero votes in the Democratic primary, and not much more than that—13%—in the Republican primary.)
Nor does this seem to be purely a function of existing partisanship, with more conservative electorates venting more dissatisfaction in the Democratic primary: the relationship between income and Biden’s vote share was more robust than that of the town’s Democratic party registration share. So, there does seem to be some class realignment at play yet to be worked out in general election voting behavior.
Working class protest votes against Biden manifested in stronger support for Phillips in these areas, and also for Haley, though the relationship between income and the Haley write-in was not as strong as with Trump.
The trend is also partially a reversal of the 2020 trends, where Biden was not any weaker in working class areas than he was in wealthier, well-educated places, albeit as a 5th place candidate that year. In working class Berlin, he got 10.9% of the vote compared to 8.4% statewide, and in Manchester, 9.6%. Meanwhile, in Hanover, Biden got 7.3%; that year, the campus overindexed most strongly for Elizabeth Warren and Pete Buttigieg (who won the town).
If Biden is facing any difficulties with younger voters, it’s not in college towns, which feel more prim and establishmentarian in their voting sensibilities as of late. And also, the war in Gaza is not registering: the organized “Ceasefire” write-in got barely more than 1,500 votes statewide, with no discernible demographic pattern of support. The evidence from polling, though, has been stronger for nonwhite and working class erosion for Biden, and we now have evidence for it in actual ballots cast.
The first official Democratic primary is South Carolina. There, it would be surprising if Biden showed any weakness with the majority Black electorate that propelled him to the nomination in 2020. But regional variation in Biden’s vote totals could hold important clues as to how his coalition has changed from 2020 to 2024.