How Jack Wins

I’m not sure that Mikie Sherrill will win NJ’s gubernatorial race. Here’s why.

Let’s do some back of the envelope math. Kamala Harris won NJ with 2,220,713 votes, roughly 400,000 less than Biden in 2020. Trump won only 1,968,215 votes in NJ in 2024, only about 70k more than in 2020. It’s not so much that Trump did significantly better than the last time he ran, so much as Harris just did so much worse.

It might be comforting to look at this and say “OK, well clearly Republicans have some sort of ceiling of votes they hit in Jersey and all Sherrill has to do is turnout those Democrats who sat on the sidelines in 2024”. I’m not convinced. Off year gubernatorial elections have lower turnout than presidential elections. In 2017, a year where there was no shortage of angry tapped in liberal voters, only 38.5% of registered voters turned out to vote in the NJ gubernatorial election, the lowest on record.

The counterpoint to this is that Democrats still won in 2017. This is true, but they also benefited from a very specific (and fundamentally different) set of circumstances. One was that the NJ Republican Party was devastated by incumbent Governor Chris Christie’s incredibly low approval ratings. Following a series of scandals, an embarrassing run for president, and fall out from his relationship with Trump, Christie had an approval rating of only 15% in 2017, the lowest on record for a NJ governor. Murphy was running against a divided Republican Party that had nominated Christie’s lieutenant governor.

The shoe is on the other foot now for Democrats. It is now the Democrats who have to defend an unpopular state incumbent: Phil Murphy. Murphy is, by no means, at Christie levels of disfavor, but it’s still a net negative: only 35% of New Jerseyans approve of him. To put that into context, that’s worse than Trump’s approval rating in the state of 41%. The idea that Murphy’s unpopularity could drag down Sherrill in the election is not crazy. It’s why Ciatterelli is constantly bringing up Murphy- if the election were to be decided simply on his job performance as governor, Ciatterelli would win in a landslide.

Yet again, there’s a counterpoint. Elections aren’t just decided by local conditions, but also by national conditions. NJ voters dislike Trump. NJ tends to vote for the governor of the opposite party than the one in the White House. Liberal voters are going to turn out in such numbers as to render local conditions meaningless. Right?

Yet again, I’m not convinced. In 2017, this happened. But in 2017, Democrats were winning the Generic Congressional Ballot by +13. In 2025, they’re winning it by +2.6. Frustration with Donald Trump is clearly not turning automatically into support for the Democratic Party, as it did during his first term. This is especially true when you consider that Trump 2 is significantly worse than Trump 1. Even still, Democrats are a slight polling error away from having virtually no lead amongst voters.

I can see this walking in my own neighborhood. I live in an incredibly liberal and affluent town within Mikie Sherrill’s district. When I walk the streets, I have only been able to count a total of six yard signs for Sherrill. If you know my town, that is incredibly scary.

Let’s loop back to those 400,000 voters that Harris lost. If you compare the election results, it’s fairly clear that the loss of voters was overwhelmingly concentrated in working class, non-white towns and districts. It might be easy to shrug this off. Off year elections already shift whiter, wealthier, and more educated than presidential elections anyway. Yeah, but that’s bad for Sherrill.

While support among white people has increased for the Democratic Party, it’s easy to forget that white people, on net, are Republican. Trump won all white voters by +12% in 2024. That’s why Democrats win elections (at least in theory) by focusing heavily on non-white voters; their basic math is that it’s ok to lose white voters by a little as long as they can win non-white voters by really high margins. That’s true in states that are really white, and that’s even more true in states like New Jersey, which are only about 50% white. NJ Democrats rely on this high turnout to balance out the off year election bonus that Republicans have; if Murphy lost a similar proportion of voters in say Hudson, Union, and Essex County as Harris did in ‘24, he would have lost in 2021.

And if Harris couldn’t turn these voters out, I struggle to believe that Sherrill can. I don’t pretend to know what Asian, Latino, or Black voters in New Jersey want or need. But I don’t think it’s crazy to question the kind of outreach Sherrill has or has not done to these voters. Without these exact voters, her path to victory in a state as diverse as New Jersey, even with a white electorate than usual, narrows. It essentially doesn’t matter if she wins 10,000 more votes in Morris County if she loses 40,000 in Newark.

Now, let’s talk about polls. Current polls (broadly) put Sherrill in the lead, roughly by around +5%. Here’s what scares me. First, if polls are as wrong on average as they were for the 2021 gubernatorial race, that would eliminate Sherrill’s lead entirely, putting both candidates at 50-50. If polls are as wrong as they were for the presidential election in NJ in 2024, that would mean a landslide Ciatterelli victory. I don’t know that they’ll be as wrong as they were in both of these elections. But the last ten years of polling have shown a persistent inability to capture Republican support in elections. It’s not crazy to imagine that polls are- to some degree- underestimating support for Ciatterelli, like they did in 2021.

But if the polls are accurate and Sherrill wins by roughly +5%, that also carries with it significant implications. That would mean, that even after a year of Trump in office, Sherrill would not improve on Harris’ margin of victory in the state whatsoever. Yes, it would be a victory. But it would hardly be a strong endorsement of the Democratic Party.