Election Eve Live Blog – 2018

*7:15pm – That’s it for today. Sleep tight. Dream of  election night surges, the hugs of loved ones, the deep exhale of knowing that tomorrow begins the long trek back to balance and sanity.

I recommend getting to bed immediately, because it’s gonna be a loooooong one tomorrow. One more sleep.


*6:34pm – Texas has seen one of the largest increases in early voting. A big question is whether or not that is a result of a surge of new voters or “cannibalization” of regular voters who are just choosing to vote early instead of on election day.

early voting.jpg


*5:00pm – Perhaps the most unusual aspect to this midterm election has been money. Unlike at any time before, the Ds have dramatically outraised the Rs in House races. If there is a surprise in the Ds direction, the money might be pointed to as the sign.

The latest:

Republicans – 43.1% of total – $652.9 mil [ Candidates- $403.6 mil ; Outside Groups- $249.3 mil]

Democrats – 56.9% of total – $851.7 mil  [Candidates- $535.6 mil Outside Groups- $325.1 mil]

In the 2016 race for the House, Rs spent $543 mil and Ds spent $423 mil


*4:45pm – Gerrymandering is designed to help a party win/keep more seats than it otherwise would based on its share of the vote. You know the idea: design districts so that your party has enough of “your kind” of voters to win, but not too many. Too many of your own voters means they are wasted. Instead, spread them out, and jam all your opponents voters into the same districts.

Makes sense.

But that means that many more of your seats are not necessarily sure things in a wave election. Your party may be able to win by about 5% in many seats every election. But if there is a wave and the other party starts over-performing by 6,7, or 8%, then there’s a sharp drop off, with a large number of endangered districts.

The term that’s often used is that a gerrymandered map is “wide but shallow.”

That’s because gerrymandering gives one party more overall seats, but fewer solid/always safe seats. So depending on the size of the D “wave” this election (if there is one), the Rs gerrymandering in certain states may actually help flip more seats that otherwise would be flipped to the Ds.


*4:03pm – Here’s a turnout model released yesterday (US Election Project) based on various factors, especial early voters. Remember, in general, Ds are rooting for highest turnout possible. But it’s not 100% clear that a higher turnout is better for them.

Summary:

Predicted 2018 turnout: 44.8% (it was 36% in 2014). From a high of 61.5% in MN and a low of 36% in KY. These are guesses, so they may be wildly off.


*3:40pm – Donald Trump essentially secured the presidency because he won over a larger chunk of voters who made up their mind in the last few days. As usually happens, these undecided voters “broke” toward him. Undecideds rarely split up evenly.  If 8% of voters say that are unsure in a race, in the final vote tally, each candidate does not usually get 4% of them. It’s more likely a 6% – 2% or 7% – 1% split. They tend to break toward one candidate.

So when you see polls that show a decent chunk of undecided voters, odds are that one candidate will get a larger share of them, and it could put them over the top if the race is close.

Some are arguing that Ds seem to be gaining (in House races) over these final days. Are undecideds breaking toward them in these last days? For example,
demsclosestrong.png


*3:20pm – Early vote figures out of Colorado:

coearly vote


*3:10pm – Early voting lines back in Champaign, IL at the Illini Union (from @MaxNWeiss). University of Illinois students are in a swing district this year, with Congressman Rodney Davis (R) at risk of losing his seat to Betsy Dirkesen Londrigen (D).

FTE gives the Ds about a 28% chance of flipping this seat. She’s an underdog, but it’s in reach in a wave year. To win, Londrigen definitely needs to rack up big margins with high turnout among students.

illini union.jpg


*3:05pm – Perhaps the most high profile race in the country is Texas Senate. Ted Cruz v. Beto O’Rourke. Beto is a sensation, raising more more than any Senate candidate ever. Presidential talk started as soon as his video on NFL players kneeling went viral. But national celebrity is not the same as winning a Texas election.

For awhile now, it was assumed that Beto had little to no chance. All talk was on whether he could prevent getting blown out, keeping a respectable national profile alive.

But, the very last poll of the race, has it tied. Is it really a toss up? No. Cruz is clearly the favorite. But in a wave D election with polls off a few points, it’s not out of the question.

From FiveThirtyEight:

texas.jpg


*3:00pm – From CNN, the weather for election day seems to be stormy  in the East:

Every state east of the Mississippi River is likely to see rain at some point when the polls are open on Tuesday, though some states will undoubtedly see more storms and potentially disruptive weather than others.

stormy.jpg

The old story is that rain is good for Rs and bad for Ds. That’s because R voters historically are more reliable, and Ds count on a higher turnout with votes from the infrequent voter.

However, there is not great data on that ever being the case. And one argument is that it helps whoever is already most motivated to vote, which this year seems to be D voters by a little bit.


*2:50pm – For the super-geeks, here’s the handiest county-by-county 2016 vote count spreadsheet that I could find (from Greg Giroux) . Very easy to navigate. Let’s you browse as comparison for this election or just to kill time. When things get extra tight on election night and the anchors talk about what countries still have outstanding votes, these are the sort of figures that let you know what to expect from those places.

For example, my current locale, Winchester, VA went for Clinton over Trump, 48% – 44%. Hometown of Kankakee Co, IL went for Trump, 53% – 41%


*2:30pm – There are a few high profile election forecast models. They use a range of factors to make their guesses. They heavily rely on polls (obviously) along with fundraising figures and other indicators.

Alternatively, there are a few “political science” forecasts that predict swings from party to party based on big historical trends. These forecasts don’t judge individual races but instead look at broad circumstances, presidential approval, general political climate, etc.

Sabato’s Crystal Ball provided a handy chart of what four of these political science models project:

Politicalsciencemodels

You’ll notice that these forecasts fit within current poll-based models, ranging from a Democratic gain of 27 to 44 House seats (all of them predict D control of House). They also project a few seat gain for Rs in the Senate.


*2:15pm – Something I should have known earlier but just learned now: Why election day is on a Tuesday in November. As you might expect, it’s because of what life was like at the nation’s founding.

November was picked because it was after harvesting but before the worst of the winter.

Tuesday was picked because it was between worship on Sundays and market days on Wednesdays. Monday was used to travel to the voting site (via horse, obviously).

Fascinating, but does it really make sense today? I’m all for traditions, yet not everything older is better.  We have extra large room for improvement in our current election process.


*2:05pm – Without a Presidential race, it’ll be easier for the “disappointed” side to spin the results tomorrow so they don’t look as bad. The most likely outcome is a D House and an R Senate. But, make no mistake, that is not a “draw.” That is a win for the Ds and would have dramatic effects on what policies could become law in the next two years. That would mark a shift in power from our current status.

Here’s a possible measuring stick for tomorrow to evaluate whether the Ds over-perform or under-perform (from FiveThirtyEight):

expectations 2


*1:45pm – Speaking of odds, the Grandmaster election forecaster, Nate Silver, constantly emphasizes that the range of possibilities is very large. In other words, don’t be too shocked if the Ds don’t win the House. BUT also don’t be too shocked if they win 55 or 60 seats (they need 23). 

silver.png


*1:25pm – A quick note on odds. It is best to think of the election models as probabilities of winning (not hard absolute guesses on who will win). There is an 85% chance the Ds take the House and an 85% chance Rs keep the Senate. But that means there is a 15% chance the opposite happens in each.

Consider field goal odds as a comparison:

An NFL kicker attempting a 37 yard field goal.

Each team gets one kick. If the Ds make theirs, they win the House. If the Rs make theirs, they win the Senate.

If you were guessing, you’d say that they both will hit. BUT, is it out of the question that one might miss? It could totally happen. You would expect the kicker to make a 37 yarder, but you wouldn’t faint if one missed. Keep that in mind whenever you hear pundits talk tomorrow about what is or is not a longshot.


*1:00pm – One of the final set of polls that we will get before tomorrow…

final polls


*12:55pm – Much ballyhoo about early voting numbers this year. It’s a very real and somewhat startling increase from the last midterm election. BUT, a dash of cold water: 2014 midterms were the lowest turnout in 70 years. That doesn’t necessarily discount the surge this year, but there was a LOT of room for improvement from 2014 anyway.


*12:46pm – A good Andrew Sullivan piece on tomorrow’s stakes:

sullivan summary_LI.jpg


*12:31pm – The ballot initiatives I’ll be watching tomorrow:

(1) Four states will vote on marijuana legalization: North Dakota and Michigan (recreational); Missouri & Utah (medical)

(2) Medicaid expansion is on the ballot in Idaho, Montana, Nebraska, & Utah.

(3) Ohio Drug Laws: All felony nonviolent drug offenses would be reduced to misdemeanors. The idea is to lower the prison population. Fearmongers claim that it will make Ohio a “drug magnet.” I think it is an interesting test point to public opinion on easing drug laws in a swing state.

(4) Louisiana Verdicts: Would ban felony convictions without unanimous jury vote. Louisiana is one of 2 states (Oregon) where you can be found guilty, even if a jury is not unanimous. This law traces back to Jim Crow; it is ridiculous that it’s even in doubt.

(5) Florida Felony Voting: A proposition would restore voting rights to felons who have completed their sentence (except those with murder or sex offenses). 1.4 million Americans may have their rights restored.

(6) Abortion-related initiatives are on the ballot in Oregon, West Virginia, and Alabama. With the new SCOUTS makeup, we can expect even more abortion drama than usual in the coming years. States will try to restrict more and more, the laws will be challenged, and those challenges will likely end up in front of the new Nine.


*11:55am Be sure to read the whole thread. This highlights something that will happen a lot over the next 36 hours. A stat pulled out of somewhere that seems good/bad for one side, but the overall impact is far more muddled.

Like in fantasy football, you can cherry pick stats or matchups to make anyone seem liable to score a lot of points. What is left out is often more important than what is mentioned. 

generic ballot.png


*11:50am Hmmm…

hispanic.png


*11:45am The final FiveTirtyEight Projections:

House: 87% chance of D control (average gain of 39 seats)

Senate: 84% chance of R control (Rs gain 1 seat)

Governor: Ds gain 8 states (63.5% of population lives in D-held state)


*11:15am More younger voters is almost always a good sign for Ds.

tx early vote.png


*11:10am We must wait until tomorrow evening for real results. Dammit. But what if we want to know right now how things are looking? We have 2 things: (1) Polls/Models ; (2) Early Voting Statistics.

If you are as obsessed with this as I am, you know all about it. But if you are a normal person, you preserved your sanity and devoted attention to other things the last few weeks. So here’s the briefest summary (more detail later):

(1) Polls/Models: Most models have the Ds picking up 25-40 seats in the House, taking control (they need 23). The trend lines over the last week (based on individual polls and generic ballot polls) has been a tick even better for the Ds. In the Senate, the opposite is the case, with most models showing the Rs picking up 1-3 seats and keeping control.

(2) Early Voting: There is significant controversy over whether early voting figures tell us anything. However, for those who believe they are helpful, the #1 storyline is that turnout is UP compared to the last midterm election virtually everywhere. Higher turnout usually is good news for Ds. However, in our hyper polarized environment, some argue that both sides are benefiting and it may not be as much of a boon for Ds as usual. For me, I think the old rule applies here, and big numbers is a cause for optimism for Ds in close races.


*11:05am There is no presidential race, so how should we judge tomorrow? That’s like a newbie sitting down as the season finale of Game of Thrones starts and asking, “So what’s going on?” 

Oh boy, where to begin…

The moon view is that tomorrow reveals how the country feels about its decision to elect Donald Trump two years ago. Was it a good idea? Have his actions/decisions been good or bad for the country?

There are 5 things I’ll be watching tomorrow:

(1) Control of the US House R’s currently control it. Ds need to flip 23 seats to take control. All 435 seats are up, but about 310 of those seats are not in doubt. Instead there are around 115 seats that are “in play.”  Almost everyone agrees that the Ds will gain some seats, but how many? Throughout the night, the biggest thing to track is the number that the Ds are seeming to flip. Is it over or under 23. Current projections: Anywhere from 15 to 60.

(2) Control of the US Senate – Rs currently control it. Ds need to flip 2 seats to take control. Only 35 seats are up, and about 12 are “in play.” More of those in play seats are held by Ds already, so they have far fewer opportunities to flip a seat. Current projections: Anywhere from Ds gaining 2 to Rs gaining 5.

(3) Governor Races – Right now there are 33 R governors and 16 D governors. But Ds are poised to make gains in both total states with a D governor and total population living under a D governor. The highest profile races are in Florida, Georgia, and Kansas.  Current projections: Ds gaining 5 to 10 seats.

(4) State/Local Races – The people who you elect to your statehouse, city council, and county board may actually impact your life more than those at the federal level. The party that wins these statehouses now will be involved in redistricting after the upcoming 2020 census.

(5) Ballot Initiatives  – Legalized marijuana, felony voting rights, and other policies will be decided by voters in states across the country.


*10:50pm – Cool map from Bloomberg showing the most common types of issues campaign ads were shown each area. (3 million ads analyzed). My own district (the top of Virginia) is the dark blue “Anti-Trump.” It makes sense, our sitting Congresswoman is a Republican, but Clinton won this district by 10 points, and we are outside DC where Trump is loathed with the fury of a thousand red giants.

This is also a good reminder that when you hear one pundit or another say that X is the issue that matters. The Democrats path to victory is Y.  They’re almost always oversimplifying to the point of uselessness. Different people care about different things. Winning control of any chamber or statehouse or school board or governor’s mansion requires different strategies.

adsmap.png


*9:54amBe sure to take care of yourself, even if your nails are bitten down to nubs.

stress drinking

stress drinking2


7:48am – Christmas Eve is often more special than the big day. It’s the same with vacations. Dreaming of what’s ahead, the building excitement can be more valuable than the live experience. Science backs that up.

vacation

The mind can work magic with unlimited potential. Enjoy this day, political/election nerds, because tomorrow might be the beginning of the hangover (or maybe not, who’s to say).


*7:35am –  Antici…

antici GIF

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s