I love it when I get to foreshadow – even if it’s not in this particular venue. The other day, on a social media post about voter turnout, I noted:
But I’ve been looking at turnout lately too. Might be a little sumthin’ sumthin’ on that subject from monoblogue in the next few days.
Here is that sumthin’ sumthin’.
Most of what you hear about polls are the topline results: i.e. “Larry Hogan leads the latest Mason-Dixon Poll by fifteen points.” Yet that may or may not reflect the reality of what is really going on because, in order to have the most accurate poll, you need the most accurate sample. The reason Larry Hogan more or less snuck up on us to become governor was that Democrat turnout in 2014 was abysmal to a point where no one predicted it would be that bad – and many of those Democrats who showed up voted for Hogan. Pollsters didn’t have that sort of turnout model in their realm of possibility, although that year’s last Gonzales Poll was closest (and closest to the result among “non-partisan” pollsters.)
As a warning up front, this post will have a LOT of numbers. But what I did was take all three recent major polls (Gonzales, Goucher College, and Mason-Dixon) and, based on their interpolated selections by party, came up with numbers that equaled 100 percent, with a small percentage added for Libertarian and Green Party candidates based on best guess of mine since they’re not polled. (That’s why their numbers don’t change much despite the varying scenarios – I used the same mix for both minor-party candidates.)
Then I came up with several alternate voter turnout universes based on the latest voter registration numbers and turnout figures provided by the state Board of Elections. The voter turnout universes I came up with were a 2014 universe (which is probably a best-case scenario for the GOP), a 2010 universe (this was the TEA Party wave), a 2006 universe (bad year for the GOP, both nationally and in Maryland), and a 2008 universe that applies the more Democrat-friendly Presidential numbers to a state election.
The fifth and final scenario was the absolute worst-case one I could think of for Larry Hogan – massive Democrat turnout on the order of 2008 combined with soft GOP and independent turnout. It’s an Indivisible dream – but does the #bluewave pan out?
This is the first time I have tried to drop a table into a post, so hopefully it’s legible. This is direct from the spreadsheet I figured out the calculations on.
2014 | 2010 | 2008 | |||||
Gonzales | Hogan | 1,076,589 | 57.7% | 1,205,595 | 56.5% | 1,709,578 | 55.4% |
Jealous | 744,174 | 39.9% | 874,778 | 41.0% | 1,302,241 | 42.2% | |
Quinn | 28,362 | 1.5% | 31,395 | 1.5% | 45,089 | 1.5% | |
Schlakman | 17,625 | 0.9% | 20,307 | 1.0% | 30,508 | 1.0% | |
Goucher | Hogan | 1,175,708 | 62.8% | 1,312,495 | 61.8% | 1,880,884 | 60.8% |
Jealous | 650,997 | 34.8% | 758,190 | 35.7% | 1,138,942 | 36.8% | |
Quinn | 28,362 | 1.5% | 31,395 | 1.5% | 45,089 | 1.5% | |
Schlakman | 17,625 | 0.9% | 20,307 | 1.0% | 30,508 | 1.0% | |
Mason Dixon | Hogan | 1,083,769 | 58.1% | 1,205,688 | 57.0% | 1,708,522 | 55.8% |
Jealous | 736,994 | 39.5% | 858,675 | 40.6% | 1,276,753 | 41.7% | |
Quinn | 28,362 | 1.5% | 31,395 | 1.5% | 45,089 | 1.5% | |
Schlakman | 17,625 | 0.9% | 20,307 | 1.0% | 30,508 | 1.0% |
2006 | worst case | ||||
Gonzales | Hogan | 1,269,353 | 55.8% | 1,336,851 | 51.5% |
Jealous | 950,092 | 41.8% | 1,200,296 | 46.2% | |
Quinn | 33,093 | 1.5% | 31,924 | 1.2% | |
Schlakman | 22,257 | 1.0% | 26,751 | 1.0% | |
Goucher | Hogan | 1,385,124 | 61.2% | 1,488,953 | 57.5% |
Jealous | 822,919 | 36.4% | 1,040,141 | 40.2% | |
Quinn | 33,093 | 1.5% | 31,924 | 1.2% | |
Schlakman | 22,257 | 1.0% | 26,751 | 1.0% | |
Mason Dixon | Hogan | 1,269,456 | 56.2% | 1,343,601 | 52.0% |
Jealous | 932,234 | 41.3% | 1,179,552 | 45.7% | |
Quinn | 33,093 | 1.5% | 31,924 | 1.2% | |
Schlakman | 22,257 | 1.0% | 26,751 | 1.0% |
Obviously the various scenarios are all there, with the worst-case being the far right-hand in the bottom page – try as I might I couldn’t get all five sets to fit in one line. In order from top to bottom, the numbers come from the Gonzales Poll (topline: Hogan +16), the Goucher College Poll (Hogan +22) and Mason-Dixon (Hogan +15).
Given these turnout scenarios, the lone path to victory for Jealous would be getting only hardcore D’s out to the polls that aren’t accounted for in these turnout models because he has to bring his numbers among Democrats to the point where they were in 2014 (Anthony Brown was in the low- to mid-70’s among D’s.) In the three polls, extrapolating numbers on a 50-50 basis from a 100% total, Hogan’s Democrat support ranges from 35 to nearly 45 percent. Unless Jealous can pull off a turnout that’s the absolute nightmare scenario (to the GOP), he’s a loser if Hogan gets even 25 percent of the Democrat vote – and Hogan’s polling beyond that even with taking away the currently undecideds. Moreover, the Kavanaugh scandal is primed to drive GOP turnout and there’s almost zero split among GOP loyalists so a nightmare scenario becomes less likely.
But more Republicans to the polls can assist in downticket races, such as the uphill battles being faced statewide by U.S. Senate hopeful Tony Campbell and Attorney General aspirant Craig Wolf as well as the “drive for five” new GOP State Senators to allow Hogan vetoes to be sustained. We’ll have to see how that goes as time goes on, but this was a fun exercise and it’s a pretty handy spreadsheet for me to keep around.
One thought on “Fun with numbers”
Comments are closed.