Skip links

A Last Word on Virginia

Dear Readers: Join us Thursday, Nov. 4 from 6:30 p.m. to 8 p.m. for our annual American Democracy Conference. The virtual event will begin with a presentation by Project Home Fire’s Larry Schack and Mick McWilliams on our ongoing polling and data analytics project. They will be followed by discussions on the state of politics moderated by Carah Whaley, Assistant Director of the James Madison Center for Civic Engagement, and featuring UVA Center for Politics resident scholars Jamelle Bouie, David Ramadan, and Tara Setmayer; former Rep. Barbara Comstock (R, VA-10); and Crystal Ball Managing Editor Kyle Kondik.

The free, virtual event will be available at this link:

— The Editors


— This Virginia gubernatorial race is one of the most vexing races we can remember.

— Terry McAuliffe (D) retains the advantage of running in a state that is clearly trending Democratic. But Glenn Youngkin (R) has many significant advantages of his own, which may outweigh Virginia’s Democratic lean.

— When assessing what the eventual results tell us about the political environment, one’s analysis shouldn’t change much if the race itself ends up being as close as the bulk of the polls project, no matter which candidate wins.

— The Virginia gubernatorial race, particularly recently, has provided clues for the upcoming midterm, but it does not always predict the future.

Table 1: Crystal Ball gubernatorial rating change

Governor Old Rating New Rating
VA Open (Northam, D) Leans Democratic Leans Republican

How Virginia has, and hasn’t, predicted the midterms

As we reach the end of the Virginia statewide campaign, we find ourselves a bit stumped about what is likelier to happen. But before we offer some concluding thoughts on the race, we thought we’d dig a little bit deeper on a question we often receive: How well does the Virginia gubernatorial race, a rare odd-numbered-year statewide race that takes place between a presidential election and a midterm, project the future? In other words, do the trends one sees in a Virginia gubernatorial race end up being replicated in the midterm?

The history is murkier than one might think, although the race has acted as a better bellwether more recently. Table 2 shows the Virginia gubernatorial races since 1965 along with what happened in the subsequent year’s midterm election.

Table 2: Virginia gubernatorial races vs. subsequent midterms, 1965-2017

When Democrat Mills Godwin won the governorship in 1965, Virginia was at the tail end of an eight-decade, one-party period of dominance, the last of 21 consecutive Democrats to hold the top job in Richmond. Naturally in this era, Virginia was not much of a national barometer, and Godwin’s comfortable victory in 1965 gave not a hint of the anti-Lyndon Johnson Republican wave of 1966 driven by a backlash to LBJ’s liberal domestic agenda, Vietnam, and more.

Four years later, Virginia finally became a two-party state, and the GOP’s Linwood Holton — a Richard Nixon ally who capitalized on Democratic factional in-fighting — handily won the governorship in 1969. Yet by the time Nixon’s 1970 midterm occurred, economic problems and more angst about Vietnam gave the Democrats additional U.S. House seats and governorships, though Republicans picked up a few in the U.S. Senate. Holton, a progressive Republican who endorsed Democrats later in his life, passed away last week.

Virginia’s 1973 result was even less representative of what would happen in the 1974 midterms. Mills Godwin switched parties and narrowly won a second term as governor even after Watergate had picked up steam. But Democrats secured a massive landslide nationally shortly after President Ford’s pardon of former President Nixon. He remains the only modern Virginian to serve non-consecutive terms as governor, although McAuliffe is trying to become the second.

At last, in 1977, Virginia held a gubernatorial election that was a harbinger of the following year. President Jimmy Carter, who had only narrowly lost Virginia in 1976, became unpopular quickly, and the GOP candidate for governor, John Dalton, won a landslide over a close Carter ally, former Lt. Gov. Henry Howell (D). (Howell had run as an independent in 1973, and the Democrats didn’t field a candidate.) In 1978, anti-Carter sentiment was more widespread, and Republicans won a modest number of additional congressional seats and governor’s chairs.

The parties switched roles four years later. In 1981, Virginia Democrats won the governorship after a 12-year drought with LBJ’s son-in-law, Chuck Robb. Subsequently, in the midst of an economic downturn, Democrats did well in President Reagan’s first set of midterm contests in 1982, although they failed to make up ground in the Senate, which the Republicans had won in 1980 for the first time in roughly a quarter century.

Virginia Democrats posted two more gubernatorial victories in 1985 and 1989. One could argue that Gerald Baliles’ win in ’85 suggested the Democratic takeover of the U.S. Senate in 1986, though that’s a bit of a stretch, and it ignores the fact Republicans did very well in the 1986 gubernatorial races (aided by the exit of many term-limited Democratic incumbents). And it is difficult to contend that Doug Wilder’s paper-thin victory in 1989 for governor said anything about 1990, which was an unmemorable status-quo midterm with minimal Democratic gains under President George H.W. Bush.

One of the Virginia elections most often cited as a harbinger is 1993, when George Allen won a landslide as governor by running against President Clinton as much as his opponent and other state Democrats. The anti-Clinton wave crested with the “Republican Revolution” on Capitol Hill in 1994, which produced the first GOP majority in the House in 40 years and a large shift toward the Republicans in both the Senate and the governorships.

Yet 4 years later, another Republican landslide for a Virginia governor foretold nothing about 1998. Jim Gilmore’s handsome 1997 victory was a misleading indicator about 1998, when GOP overreach on President Clinton’s impeachment robbed Republicans of a prime opportunity to build up a large Senate margin. At the same time, the GOP actually lost a few House seats — and this was Clinton’s “sixth year itch” election, when his party ought to have been vulnerable.

It was turnabout for the Democrats in the next Virginia cycle. Mark Warner (D) won the governorship in 2001, but post-Sept. 11 Republicans fared well in President George W. Bush’s first midterm. Bush’s sky-high popularity made some difference in ’02 — though it had had no impact on the ’01 Virginia race. Federal and state contests can run on very different train tracks, which is another reason why comparisons between a state’s gubernatorial outcome and a national midterm can be tricky. Democrats did net a few governorships in 2002 despite the good GOP performance in Congress.

In recent years, the Virginia results have more often offered a preview of the future, fueling the belief that there is a straight-line connection between this off-year election and the next year’s midterm.

Tim Kaine (D) succeeded Mark Warner as governor in 2005, and the Democrats took over both houses of Congress in 2006. President Bush is one of the vital links, of course. He was already down in the polls by November 2005 because of the Iraq War and his mishandling of Hurricane Katrina, and it only got worse for Bush in 2006.

In 2009, Bob McDonnell (R) scored a massive 17 percentage point knockout for governor, partly because President Obama’s ratings were sinking. The health care battle and the balky economy energized Republicans and discouraged Democratic turnout (though a weak Virginia Democratic nominee played a major role as well). One could contend that McDonnell’s triumph presaged Scott Brown’s January 2010 capture of Ted Kennedy’s old Senate seat in Massachusetts and the national Republican takeover of the House in November 2010.

Eight years ago, in 2013, McAuliffe (D) won the governorship by just 2.5 points over then-state Attorney General Ken Cuccinelli (R), a margin that was closer than the polls suggested. McAuliffe’s victory — the first time since 1973 that the presidential party candidate won the governorship — did not predict Democratic gains nationally, as Republicans built their biggest House majority since before the Great Depression, flipped control of the Senate, and netted a few governorships.

And most recently, in 2017, Ralph Northam (D) ran ahead of the polls and beat Ed Gillespie (R) by about 9 points. That did reflect a Democratic-leaning environment nationally, particularly in diverse and highly-educated suburban areas like Northern Virginia. That environment continued into 2018, when Democrats flipped the House and netted several governorships, though they did not win back the Senate in large part due to a very challenging map.

In summary, 3 of the last 4 Virginia elections did end up being bellwethers for the future, and perhaps the surprising closeness of McAuliffe’s 2013 victory ended up being something of harbinger for 2014 as well. The Virginia results from the more distant past are more mixed, with some elections serving as previews of the following year’s midterm and others not.

Closing thoughts on Virginia

Whatever happens on Tuesday, the takeaways for the national environment shouldn’t be all that different whether McAuliffe wins by a tiny margin or Youngkin wins by a similarly small margin.

The more interesting result would be if the race broke clearly one way or the other — as in, if either candidate won by more than a few points. A Youngkin win by several points would offer confirmation that the political environment has broken wide open against Democrats. Meanwhile, if McAuliffe wins by several points, it may indicate that Biden’s poor approval ratings are not as much of a drag on Democrats as one might otherwise think. There will be plenty of time to analyze the results after we get them.

But let’s assume that the bulk of the polls are correct and that the race ends up being very close either way. Both of those possible results — a narrow win by either candidate — would suggest a significant falloff for Democrats from their strong Virginia performances in the Trump era and represent, at the very least, a bright red “check engine” light at the midpoint of the Democrats’ journey from last year’s presidential race to next year’s midterm.

Now… as for the race itself.

We know based on President Biden’s sagging approval rating that the environment is, frankly, horrible for Democrats. As Washington Democrats have dithered over parallel negotiations to pass a bipartisan infrastructure bill and a Democrats-only social spending package, they have seen numerous other problems mount. The messy withdrawal from Afghanistan dented perceptions of Biden’s competence, and he has not recovered. Other problems, like inflation, supply-chain problems, and gas prices — along with the lingering pandemic — have likely chipped away at Biden and the Democrats as well. Biden’s national approval rating is in the low-to-mid 40s, meaning that his approval is lagging several points behind his 51% of the national popular vote last year. The dropoff appears more stark in state-level polling. Biden got 54% in Virginia, but his approval rating in several polls is often 10 points or more worse than that. We’ve seen similar Biden drop-offs in other recently-released polls in states bluer than Virginia, such as Connecticut, Maryland, and New Jersey.

McAuliffe will need some Biden disapprovers to win — there may be some disaffected Democrats who nonetheless will vote for McAuliffe, but in a nationalized era, having to win presidential disapprovers is difficult for any candidate. Democrats hope that the polls are underestimating them, as the polls did in the recent California gubernatorial recall. But the trendlines in the Virginia polls have been bad for the Democrats, as several pollsters who have conducted multiple polls of the race have shown earlier McAuliffe leads disappearing in the closing stages.

We have also previously noted the history in Virginia — the party that holds the White House has, for decades, performed worse in the governor’s race than they did in the previous year’s presidential race. The level of dropoff varies significantly, but the reality of the trend does not. And while elite opinion can be just as wrong, or worse, than polls, we have been struck by the number of our best sources who have told us that they believe Youngkin will win — even some Democrats, though there naturally was a reluctance to admit it. It is worth remembering that there also was widespread elite pessimism about Northam 4 years ago — pessimism that ended up being wildly misplaced — although the sentiment toward Youngkin is better supported by the numbers and the environment this time.

We have had this race rated Leans Democratic ever since March, when we first issued our ratings. At the time, we were criticized by some for not having the race rated at least Likely Democratic — such was the conventional wisdom that the race was a shoo-in for Democrats. But the history gave us pause, and we kept the rating after Republicans nominated Youngkin, who is conservative but not as conservative (and less Trumpy) than some of the alternatives were.

The race has been very close for a long time, based on our understanding of both public and private polling and modeling. We kept the race at Leans Democratic anyway because, even in a close race, we’ve come to believe that a state’s baseline partisanship can break ties in favor of the state’s stronger party — which is the Democrats in Virginia.

But as we reach the end here, we have to say: We do not really have a strong handicap either way in this race. There are some indications that suggest this could be the closest Virginia gubernatorial race since 1989, when Doug Wilder (D) won by less than half a percentage point. The polls, taken collectively, don’t strongly point to an obvious winner, nor do other factors, like the more than 1.1 million votes cast before Election Day. Those votes will be Democratic — it’s just a question of how Democratic. The McAuliffe campaign claims an edge in the neighborhood of 60%-40% and that their side had a strong finish over the weekend, although we have heard of other models of the votes already cast that show a smaller Democratic advantage. However, the majority of the vote will come on Election Day itself if the turnout of registered voters is in the same ballpark as the past couple of gubernatorial races, and this vote will be substantially more Republican. Youngkin has the enthusiasm, the environment, the history, and perhaps even the issues (given his focus on education and its increasing salience in polling). McAuliffe has the state’s Democratic lean in his favor.

However, we do feel we owe it to readers to push this race one way or the other and not just move it to a Toss-up rating at the end. So we’re moving from Leans Democratic to Leans Republican. Our sense is that the race has been moving toward Youngkin, in large part because of the political environment. McAuliffe’s Trump-centric campaign also just doesn’t seem as potent in a non-federal race with the former president no longer in the White House.

Those who believe McAuliffe will still win — and they very well may be right — will remember 4 years ago, when many seemed to think Ed Gillespie (R) had controlled the narrative of the race and was headed for an upset. But the difference between now and then is that Ralph Northam (D) had the advantages of voter enthusiasm, history, and political environment. As noted above, Youngkin benefits from those factors now. That may be enough to get him across the finish line, although a victory by either candidate remains on the table.

A couple of additional things to watch tomorrow: Because of changes to Virginia law, mail-in and early ballots will be tabulated more quickly than in 2020, so instead of the Republicans jumping out to an early lead, Democrats are likely to. Then it will be a question of whether Youngkin catches up to and passes McAuliffe, and there may be late drama depending on the rhythm of the vote count. Also, for all of the focus on the suburbs, keep an eye on the red, rural counties in central and western Virginia. The turnout and the GOP performance in those areas are a very important part of this story too. So too is Democratic turnout and performance in heavily Black areas, particularly in the Greater Richmond and Hampton Roads regions. In an election that has the potential to be extremely close, it’s never just one place or one group that decides the outcome.

Note: This article contains some material adapted from a Crystal Ball story published in 2013.