Republicans in California have emphasized VBM for years. It has its inherent advantages. First and foremost, it’s a vote in you pocket rather than one you might get later. It also means you have to send out one less mailer, make one less phone call, and one less door knock. Or more because you’ll hit that one voter repeatedly until election day. As I’ll show below VBM voters show up in greater numbers.
This meant that Republicans would do very well in vote by mail, but crash and burn on election day. In 2010, VBM was 43%D/39%R, but the final vote was 45%D/34%R. In 2012, VBM was 43%D/36%R and the final vote was 45%D/31%R.
In 2010, there were 5.7 voters VBM and 11.5 precinct. In the final vote, 87% of VBM voters sent in their ballot, but only 46% of precinct voters showed up. In 2012, VBM voters had increased to 7.8 million. This time 86% of VBM voters sent in their ballots and 62% of precinct voters showed up. Since there were more VBM voters in 2012, the total VBM voters rose from 5 million to 6.7 million.
Democrats were winning elections but it was harder for them than it needed to be. They decided to encourage their voters to VBM and were going to emphasize getting those votes in early. They started with the 2014 primary. And it worked. Democrats got 96% of their 2012 pre-election day VBM voters and Republicans only got 91% of theirs. So while Democratic VBMs were 5.5% greater than Republicans in 2012, they were 7.6% greater in 2014.
Win for the Democrats? Well, sort of. While I don’t have overall turn out for either primary I do have the 2014 results. They were a mixed bag. Democrats did better in the final vote than 2012 in some races and worse in others. They generally did better in races where there was a Democratic incumbent in 2014 but none in 2012 and those where Democratic candidates dramatically outspent Republican candidates. Overall, Democrats didn’t do better and they did lose out on top two in a possible competitive district, CA-25. But it was a win because they achieved their objective of getting the votes in early. Of course the other objective of doing better in the primary was elusive. The Democratic emphasis on early VBM voting changed the ballot dynamic. Whereas before, Republicans did better in VBM than election day 90% of the time, they only did better on election in 20% of races both parties were competing in. It was a remarkable change. Was it a fluke? Perhaps. But it may have been Democrats moving their precinct voters to VBM. And really who cares when you get the vote as long as you get the vote?
That brings us to the 2014 general election. The total VBM ballots are, at this point, 1% lower than they were in 2010. This wouldn’t be a big deal, but there are 3.4 million more VBM ballots out. So while this was 42.9% of all VBM ballots then, it’s 27.5% this year. That’s a dramatic drop. Some VBM ballots will arrive on Monday and a huge amount will be dropped off at the precincts on election day. I’m making an assumption here, but if VBMs are 1% lower than 2010, the total final number should be similar. That’d mean there’d be roughly 5 million total VBM ballots, the same as 2010. The problem is that while that was 87% of all VBM ballots, this year it’d only be 56%.
There were 11.5 million precinct voters in 2010 and they showed up at a 46% rate. This year there are 8.9 million precinct voters. If they show up at a 46% rate, we’ll have 4.1 million precinct voters. So overall we’ll have 9.1 million voters, a 12% drop from 2012.
Here’s assumption #2. If VBM is 36% lower than 2010, it makes sense that precinct voters will be 36% lower too. It may be more. Pre-election day VBM voters were only off 4% from 2012 to 2014. Election day VBM drop-offs were 23% lower and precinct voting was 26% lower. If election day drops off 36%, the rate precinct voters will show up will be 29.5% instead of 46% in 2010. That’d mean only 2.6 million precinct voters and a total of 7.6 million ballots. Or even lower if there’s a steeper drop off.
In 2014, Democrats have a 4% advantage in early VBM ballots, roughly the same as they had in 2010. In 2012, it was 8%. In 2010, Democrats had the election day voters to bail them out. Roughly 72% of all ballots were precinct voters and VBM drop offs on election day. In 2012, 68% of all ballots were precinct voters and VBM drops offs. If there is the same turnout on election day and VBMs on election day, those will account for 73% of all ballots. If there’s a drop off, they’ll account for 61% or less. That’d mean there’d be less ballots coming in to offset the current mediocre Democratic VBM performance and they wouldn’t change the results as much as either 2010 or 2012.
Of course that assumes that election day will be a lot more Democratic than VBMs, just as it was in 2010 and 2012. If it’s like the primary, it’ll be less Democratic. That means that like the primary, the current Republican advantage will grow.
I’m making a few assumptions here based on previous voting patterns. The first is that a lower percentage turnout in pre-election VBMs means a lower overall turnout, something that’ll hurt Democrats. The second is that election day might not be the panacea for Democrats it was in 2010 and 2012. It could be like the primary.
(We should note that one big election day boost Democrats got in 2012 was a slew of new registrants courtesy of online registration. That didn’t happen this year.)
So I’m left with two possibilities here. The first is to ignore the primary election/assume precinct turnout won’t drop and Democrats improve similar to 2012. The second is that this is that turnout will drop significantly and that Democrats won’t be saved by precinct voting. The first scenario should produce similar or slightly more Republican results as 2012. The second scenario should produce much more Republican results than 2012. I’ve shown here that the second scenario is more likely, but I’m also a Republican and I have a bias toward the theory Republicans will do better with.
So I’ll present you with predictions based on the two turnouts, a high one and a low one. If you think I’m hedging my bets, I understand. But if my low turnout theory is right, I’ll have predictions here no one else has.