Last night, Democrats scored some significant wins in Virginia’s House of Delegates. Some of the best handicappers said that Democratic control of the chamber was highly unlikely, but once every race is called, Democrats may end up taking the chamber.
So why did the blue team do so well in these races? Did they simply ride Ralph Northam’s coattails to victory? Or was something else happening?
As is often the case in politics, the data is limited here. But looking at results in other states and using historical data can provide some clues.
Before getting into the specifics of Virginia, it’s important to remember that not all gubernatorial wins are accompanied by similar routes in the state legislature. New Jersey Gov. Chris Christie won two terms in 2009 and 2013, but barely made a dent in Democratic legislative numbers. Republican Gov. Phil Scott of Vermont won by almost 9 points but didn’t bring Republican majorities to the state legislature with him. West Virginia Gov. Jim Justice was elected as a Democrat in 2016 (he’s since changed his affiliation to Republican) and Montana Gov. Steve Bullock (also a Democrat) both faced closer elections than Christie did in 2013 or Scott did in 2016. But both of their state legislatures remained in Republican hands.
That’s not to say there’s never a relationship between state legislative and gubernatorial results. But state legislative results often follow broad national patterns while gubernatorial elections can be more hit-or-miss.
I don’t yet have the necessary data to do an apples-to-apples comparison of Northam’s performance in each district to each House Democrat’s performance. But the data I do have suggests that the race for the Virginia House of Delegates had a very strong national component.
I ran a simple linear model that used incumbency and Hillary Clinton’s share of the 2016 presidential vote in each House district (helpfully compiled by Daily Kos Elections) to predict each Democratic House candidate’s share of the vote. This simple model doesn’t capture every important aspect of a race (candidate quality, demographics, fundraising, etc) but it performs well and explains nearly all the variation in the data. Other analysts have attacked this data in slightly different ways, but the central insight is the same–that the battle lines in the 2017 Virginia House race strongly resembled those of the 2016 presidential election.
Moreover, in many districts, the Democratic candidate’s margin didn’t differ much from Hillary Clinton’s. That doesn’t make Democrats’ performance last night unimpressive. Many Virginia Democrats were facing incumbents (who, according to my model, gained votes when all else was held equal) and Clinton was not. And Democrats contested significantly more districts than usual, which suggests that they might be able to organize and turn out effectively in 2018.
Again, the sum of the evidence from last night indicates that the national environment is currently favorable for Democrats. But, if you’ve been watching the polls and other indicators, that environment shouldn’t be a surprise.
