For election commentators, the year following a presidential contest is typically one of slim pickings. There’s a governor’s election in Virginia and one in New Jersey, and that’s about it, as far as high-profile races go (though this year there was a bonus California election).
That hasn’t traditionally stopped pundits from drawing big, broad lessons about what election results in Virginia or New Jersey might mean for national politics. The 2005 Democratic wins sent “a powerful message that President Bush’s political standing has fallen,” wrote the New York Times. The 2009 races were a “test” for Obama, and the Democratic candidates’ defeats were “humiliating” and “an unmistakable rebuke,” per Politico.
In both years, these outcomes were indeed followed by a rough midterm performance for the president’s party. Yet relatively few people in these states tended to say they’re voting to rebuke the president. For instance, in 2009, exit polls showed voters in Virginia and New Jersey continued to strongly support President Obama, even though they voted for Republicans for governor. And the candidates themselves generally don’t shape their messaging around the incumbent president.
The overall pattern, though, is tough to miss: The incumbent president’s party has, in recent decades, almost always lost these Virginia and New Jersey races.
Of the 16 governor’s elections in these two states since 1989, the incumbent president’s party has lost 15. (The sole exception was Virginia’s 2013 governor’s election, which current Democratic gubernatorial candidate Terry McAuliffe won during Barack Obama’s second term.) That fits with the general trend in which the president’s party does poorly in midterms.
Still, within the overall tendency toward backlash, there’s a fair amount of variation in just how badly they perform and how these individual races turn out. And it’s always possible that this time will be different. This time around, Democrats hope to defy the trend, and polls show they might.
But we shouldn’t necessarily get carried away with what that outcome might mean. It’s not that the Virginia and New Jersey races are irrelevant to how next year will go. It’s that each is just one part of a larger picture — with a year remaining in which the political situation could change.
Polls show a tight race in Virginia and a bigger Democratic lead in New Jersey
The Virginia contest is the closer one in the polls. The state’s former governor, Democrat Terry McAuliffe, a longtime close ally of the Clinton family, is running for another term in the office against Republican Glenn Youngkin, a wealthy former private equity executive. (Because Virginia governors can’t serve consecutive terms, the current governor, Democrat Ralph Northam, can’t run again.)
Virginia has a history of close governor’s races, but the state has gotten increasingly blue on the presidential level, with Biden beating Trump there by 10 percentage points. Polls show a tight race, with a slight edge for McAuliffe on average.
In New Jersey, incumbent Gov. Phil Murphy (D) is running for a second term against former state assemblyman Jack Ciattarelli (R). Nationally, New Jersey has been a safe state for Democrats since the 1990s, but Republican Chris Christie recently managed to win two terms before being dragged down by scandal. On average, polls have shown Murphy with a bigger lead, but there have been a few suggesting a close contest.
On the surface, governor’s races tend to be about how things are going in the state. But they can be affected by broader national trends — pandemic policies and the economy are looming large in both of these races. They can also play into national media narratives — Youngkin is attacking the purported use of “critical race theory” in schools. Meanwhile, McAuliffe is trying to tie Youngkin to Donald Trump, and Murphy is trying to do the same to Ciattarelli.
Are Virginia and New Jersey bellwethers?
Though Virginia and New Jersey have tended to swing back and forth between the parties for governor, they’ve done much less of that on the presidential level. Virginia was a solidly Republican state in presidential contests from 1968 to 2004 but has gotten bluer ever since. New Jersey, meanwhile, has voted for every Democratic presidential candidate from 1992 onward.
Still, the results do fit the general pattern of midterm backlash that’s long been common in US politics. The president’s party almost always loses seats in the House of Representatives (they did so in 17 of the 19 midterms since World War II). And that party tends to suffer in governor’s races too — they lost governor’s seats, on net, in 16 of 19 midterms in that same span. For whatever reason, when a president’s party is in office, voters seem more likely to give the other party’s candidates a shot in the midterms.
In that sense, the Virginia and New Jersey results seem to qualify as “early midterms.” But that doesn’t mean they will predict the midterm results the following year. One or two contests don’t have such totemic power. The two most unusual recent midterms — 1998 and 2002 — were essentially draws for the president’s party, which qualifies as an unusually good result for them. They weren’t really predicted by the Virginia and New Jersey races one year prior, which followed the typical pattern.
But sometimes commentators’ takes do happen to be right. After Tim Kaine kept the Virginia governor’s office in Democrats’ hands in 2005, Democrats really did continue to gain in the state — they won Virginia Senate races there in 2006 and 2008, and Obama became the first Democratic presidential candidate to win there since LBJ. Republican Bob McDonnell’s win in the 2009 governor’s race, though, did not presage a durable return of the state to the Republican fold, since Democrats have won every statewide contest there since.
Another complication is that voter behavior in state races has become increasingly nationalized, with ticket-splitting on the decline and national-level partisanship becoming more determinative of who voters support on down-ballot races. This trend is clearest in federal politics: In 2000, there were 30 senators representing states the other party’s presidential candidate won, and now there are six.
Governor’s races have not become quite as nationalized as that, but they have become more likely to match the presidential result. After the 2002 elections, there were 20 governors representing states the opposing party’s presidential candidate won. Now, there are 10. (Four are Democrats, and six are Republicans.)
Currently, Virginia and New Jersey are considered solidly Democratic states on the presidential level. Both were willing to elect Republicans as governor not too long ago. But if more voters are sticking with their presidential party no matter what, Republicans will have a far tougher time winning statewide — which means any limited usefulness these two states had as bellwethers may have declined.
That’s not to say this November’s results will tell us nothing about the national political situation. It’s fair to say that, if Republican wins materialize in these increasingly blue states, that’s not a great sign for Democrats. A close outcome will be tougher to interpret. If Terry McAuliffe wins by 2 percentage points in Virginia, is that bad for Democrats considering it’s now a blue state? Or is it what we’d expect, since that’s about how much McAuliffe won by the last time he ran, in 2013?
When trying to discern what will happen next year, it’s important to look at the whole picture rather than over-extrapolating about one or two races. For instance, there was, unusually, another high-profile governor’s race this year already: California’s recall election. There, Gov. Gavin Newsom got the exact same share of the vote that he did in 2018. Since 2018 was a strong year for Democrats, California was a good result for the party. There are also more ominous signs for Democrats, though, such as President Biden’s declining approval rating.
The news in the following year could get better for them (if the pandemic and economic situations improve) — or worse. Virginia and New Jersey will be interesting data points, but the full story hasn’t been told yet.