Monday, November 6, 2017

Election Night Preview: What to look for in Virginia and New Jersey

Here’s a quick overview on what harbingers to pay attention to as the results start rolling in tomorrow night.

Virginia
The Virginia race for governor has been competitive from the start, despite the fact that the polling has been all over the place – ranging from a 17 point Democratic advantage to an 8 point Republican edge in various polls released over the past two weeks alone. Ralph Northam seemed to have a small and consistent advantage heading into the fall, but his lead was never a comfortable one. Monmouth’s polling showed him doing relatively well in traditionally conservative parts of the commonwealth in September. That all changed as Republican Ed Gillespie focused on an anti-immigration message and the race took a decidedly nasty turn. A majority of 56% of voters described the campaign as being a largely positive affair back in late September, but that number went down to 25% just six weeks later.

Basically, Gillespie’s strategy won back his conservative base in Western Virginia, but simultaneously pushed moderate Northern Virginia voters into Northam’s camp. This means the race is going to come down to base turnout with just a few swing districts holding the key. Since Northam’s support has grown stronger in the DC suburbs, Gillespie will need to surpass his 2014 U.S. Senate performance in the western region. Our polling suggests he might just do that.  However, this still wouldn’t determine the outcome.

From Northam’s perspective, he will have to romp in Northern Virginia and pull big numbers from the Hampton Roads region. Specifically, he will need two thirds of the vote in Norfolk, Portsmouth, and Hampton County, and he must keep the margin close in Virginia Beach. If Northam exceeds these targets he will likely be the next governor. If he falls significantly behind these targets, Gillespie should emerge victorious. But if Northam is just meeting these targets, we need to look for other tea leaves to read.

The counties just north of the Greater Richmond Area have been a fairly good indicator of the commonwealth’s mood in past elections, especially around the upper Rappahannock River. If I had to pick one set of returns to watch on election night, it would be the numbers from Caroline County.

Caroline County tends to vote Democratic, but has swung to Republicans on occasion. Importantly, it has voted with the winner in every Virginia election for governor, U.S. senate, and president from 2001 to 2014. It broke this trend in 2016, giving Donald trump a 5 point margin while Hillary Clinton won the commonwealth by 5 points. However, the county has been uncannily reliable in recent gubernatorial races: giving Democrat Terry McAuliffe a 5 point edge in 2013 when he won Virginia by just over 2 points, giving Republican Bob McDonnell a 13 point edge in 2009 when he won Virginia by 17 points, and giving Democrat Tim Kaine a 10 point edge in 2005 when he won Virginia by 6 points. It was a little more bullish on Democrat Mark Warner in 2001, giving him a 22 point margin when he only won the commonwealth by 5 points that year with an electorate that looks notably different than Virginia does today.

So keep an eye on Caroline County. It has a history of voting slightly more Democratic than the rest of Virginia in every election, but broke with that streak to back Trump last year. If Northam wins this county by at least 5 points, there’s a good chance he is meeting his targets elsewhere in the commonwealth.

New Jersey
The polls have been exceedingly static in New Jersey’s race for governor, landing somewhere between a 14 to 16 point lead for Democrat Phil Murphy. Monmouth’s polling indicates that this will be a record low turnout election (*see note). Even though this means the electorate will be comprised of people who vote in nearly every election – a majority of these habitual voters say they really don’t know where either candidate stands politically.  They are simply pulling the lever for the Democrat or the Republican. And in New Jersey, that means a natural 12 point advantage for the Democrat.

Republican Kim Guadagno has done everything in her power to distance herself from Chris Christie – who is in part responsible for the GOP’s poor standing in the Garden State – but she will need to take a page out of the incumbent’s playbook if she is going to pull off a shocker.

Jon Corzine won the 2005 election by just over 10 points, but he lost re-election to Christie four years later on a 14 point swing to the Republican. This shift was fairly uniform in most of New Jersey’s 21 counties – between 8 and 14 points. But there were three counties where Christie’s performance was staggeringly good. He swung Ocean County by 23 points – going from a +12 GOP advantage in 2005 to +38 in 2009 – as well as Monmouth County by 23 points – going from a +8 to +31 margin. He also swung the Democratic bastion of Middlesex County from a 17 point deficit for the Republican nominee in 2005 to a +2 victory in 2009.

Guadagno needs to follow the same path if she is to win – i.e. put up monster numbers in large Republican counties (Ocean, Monmouth, Morris) and win at least one sizable Democratic county.  Another option would be to padlock every polling place in Hudson County and then put voting booths on the back of pickup trucks to personally visit every registered voter in the rural counties of Hunterdon, Warren, and Sussex. The fact that either scenario is about as likely to happen is pretty much all you need to know about this race.

-----------------------------------------------------------------------------

* Note on interpreting turnout trends: You cannot compare recent turnout as a percentage of registered voters to elections prior to 1997.  The Motor Voter law that went into effect in 1996 significantly increased the voter rolls in New Jersey and Virginia. While the law may have brought some new voters to the polls, it also added a lot of people to the voter rolls who never had any intention of voting. As such, turnout figures for elections prior to 1996 are higher in part because a smaller number of eligible voters were actually registered. For example, even though turnout in New Jersey’s gubernatorial elections seemed to take a massive hit from 65% of registered voters in 1993 to 56% in 1997, the decline is much less precipitous if the pool of all eligible voters is used as the base – taking turnout from 47% to 45% over that period. This doesn’t discount the fact that turnout has continued to decline, though. Since 1997, New Jersey’s gubernatorial turnout has consistently declined, hitting a record low 40% of registered voters in 2013. Virginia’s lowest gubernatorial turnout was 40% in 2009, although it rebounded to 43% four years later. [Also, see note at bottom of Virginia’s election page.]