Midterms in Review

 I went to bed on Election Day thinking that Donald Trump had a pretty good night. I knew that the Democrats had won back the House of Representatives, but that was the bare minimum for what they needed to accomplish. That majority looked somewhat modest. More importantly, they were on track to lose four senate seats that they were defending without flipping any Republican seats. When I went to bed, there was even a chance that Scott Walker migth hang on for a third term as governor of Wisconsin.

 When morning came, I learned that Walker was out and that the Democrats had booted Dean Heller out of his senate seat from Nevada. And there was suddenly more than a little doubt about the outcomes in both Arizona and Florida. In the five days since then, it seems almost certain that the Arizona senate seat will be won by the Democrats. And Florida is not done counting yet, although I suspect that Bill Nelson will come up short. For purposes of this post, we'll assume that Democrats lost a net of two seats.

The other piece of happy news since Tuesday has been the steady drip of house races going blue. There are still 13 seats that have not been called by news outlets. Nate Silver estimates that ten of those 13 seats will wind up in the Democratic column. That will mean a net gain of 38 House seats.

First, A Word About Me.

 My official prediction was that the Democrats would gain 32 seats in the house. That was slightly pessimistic. The real number will be closer to 40. That's an impressive number by an reasonable measure. The Democratic majority will be about the same size that the Republican majority has been the last 2 years. It will be awhile before we have the final results but it appears that the Democrats have won most of the very close races, which means they put their resources in the right places in the right quanities. That is cause for optimism.

 My official senate prediction was a loss of one seat, but I broke that down into four categories of races.  The below passage is from the preidcitions post, which are in the maroon background. I've also added post-elections commentary in ALL CAPS BOLD. (So you know I'm serious.)

A.  One Likely Loss. The Dems will lose North Dakota. CORRECT

B.  Two True Toss-ups. The Dems will split Missouri and Florida. CLOSE BUT NO CIGAR, UNLESS NELSON PICKS UP 12,000 VOTES IN THE RECOUNT.

C. Three Lean Democrats.  The Dems will take 2 out of 3 among Indiana, Arizona and Nevada. CORRECT

D. Long Shots  The Dems could lose Montana, but I don't think they will. They could put an upset in Tennessee but I think they will come about 3 points shy there. Texas is hard for me to judge because I so badly want Ted Cruz to lose. CORRECT, ALTHOUGH TENNESSE WAS NOT AS CLOSE AS I HOPED. BREDESEN LOST BY 11, NOT THREE.

I came pretty close to getting this exactly right. Whatever Nelson's margin of defeat winds up being will also be the difference between me and a perfect night in the upper house. (Yes, I'm aware my model built in some fudge factor by grouping races. The truth is that I overestimated the Dems in Indiana, Tennesse and Missouri more than I did in Florida, but hey results are what matters.)

I also predicted that the Democrats would gain eight governor's mansions but I didn't specify which ones. It looks like the final number will be seven. I did think Gillum and Abrahms each had a chance to win, so I guess we can put the difference there. (Full disclosure: I spent less than 90 seconds working on my Governor predictions.)

I'll gve my self an A- on the predictions. And shake my fist at the design of the Broward County ballot for keeping me from having a truly impressive night.

Report Card Time.

In my election preview I created a somewhat vague grading scheme for how to evaluate the Democrats performance on the night. (I reserved F Scenarios for failing to win either house. We can skip that, thankfully.)

D Scenarios.  If the Democrats win one house but underperform in the other, then they have done the bare minimum to stay relevant to government. The most likely D scenario is a gain of 25 seats, giving them a razor-thin margin in the House, but they lose several of their senate seats and don't pick up any GOP seats.  So about 221 in the house and 46 or 47 in the Senate. Awful, but not technically a complete failure.  WE WELL EXCEEDED THAT IN THE HOUSE, BUT WILL PROBALY END UP WITH THAT TALLY IN THE SENATE.

C Scenarios.  If the Dems win at least 30 seats and don't lose more than one senate seat (net), they will get a gentleman's C. (C+ if they do well in the governor's race.)  TECHNICALLY WE'RE IN C+ COUNTRY AS OF NOW, BUT I THINK I MAY HAVE PUT TOO MUCH VALUE ON THE SENATE FOR MY GRADING SYSTEM, ESPECIALLY GIVEN THE TOUGH TERRAIN FOR DEMOCRATS. THIS NIGHT FEELS MUCH BETTER THAN A C+, EVEN WITH THE DISSAPOINTMENTS IN FLORIDA.

B Scenarios.  A lot of the advanced analytics types have the most likely outcome for the Dems being about +37 in the House and -1 in the senate.  I think that's a solid B. THIS IS ABOUT RIGHT, BUT THE GOVERNOR'S RACES WERE ALSO STRONG. 

A Scenarios.  So what makes a wave? The biggest recent was the 2010 Tea Party election when the Dems lost 63 seats. But they started that year with 256. The GOP only has 241 at the moment. So +48 would match that end result. I think anything over 40 would be a wave and it probably means picking up at least one senate seat.  So let's call this 235 in the House and a 50/50 Senate.  WE DIDN'T GET THERE BUT THE HOUSE CAME AWFULLY CLOSE.

A+ Scenario:  This is easy-the House by more than a handful of seats and the Senate. But to really earn the +, we'd have to see some really awful Republicans lose. It's not an A+night if Ted Cruz is re-elected. Beating Steve King or Dan Rohrbacher would be nice cherries on top too. Stace Abrams winning in Georgia would also be a profound outcome for the party  WELL, DANA ROHBACHER IS A NICE PLET TO PUT ON THE MANTLE.

I'm torn between giving them a B or a B+. For now I'll call it a B but I will run back here to upgrade if Nelson and/or Gillum winds up winning. And there is going to be a runoff race in Mississippi. Hopefully we'll make that competitive at least. It sure feels like our base is energized by the results.


What Does it Mean?

The Democrats will be able to pass all kinds of stuff in the house. But if they want any of that stuff to become law, they will need to reach agreement with the senate. And it's not just a matter of persuading four Republicans to join them. Mitch McConnell will control the senate and he alone will decide what gets put to a vote. I think that means a lot of gridlock.

But there are a handful of Republican senators who are vulnerable to losing in 2020.  Specifically Susan Collins (ME), Joni Ernst (IA) and Cory Gardner (CO) will all be in very close races. There will also be an election for the remainder of John McCain's term in Arizona. John Kyl has said he will not run in that race and he will probably resign in January, allowing Arizona to nominate a person who will run. Whoever is chosen, will have to worry about facing a close election in 2020.

Those Republicans will want to run on some accomplishments. They can put pressure on McConnel to let some things get done, especially if it means avoiding a government shut down. The senate is also close enough that there will be at least some very modest limits on who Trump can appoint to the senate. (But I'm not optimistic on this point. It's pretty easy to imagine Mitch getting 50 of his 53 member coalition to vote for just about anyone.)


What We Really Care About: 2020.
The map at the top of this post was put up by Nate Silver on Twitter the day after the elections. He tabulated the "popular" vote of every house race, by state. (This required an adjustment for Florida where votes in uncontested races do not get counted.) The result was a near replic of Obama's 2012 victory. (Republicans winning Ohio and the 2nd district of Maine, for a 324-214 Democratic win.)

That's certainly encouraging. And I think that's a possible outcome for 2020, although something would have to change for Iowa to flip back to blue, having gone for Trump by eight points last time. Trump's tariffs might just be disastrous enough to swin the race, but if that's the case, Democrats have probably sewn up the race in other places.

I think the biggest take-aways from Tuesday are this:
1. Trump is not imploding but he's also no invincible. The results in senate races correlated with Trump's current popularity. The Republican candidates in Indian, Tennesse and Missouri benefitted from his popularity. Those in Nevada and Arizona suffered. And Florida, as always, was a wash.

2. The next race will be close. Trump will start 2020 with the command of his party and a solid 200 electoral votes in his pocket. The race is going to be decided in about eight states.  A few analysts have identified two paths for the Democrats to win the White House back-a northern strategy of winning back Wisonsin, Pennsylvania and Michigan or a southern strategy of winning Florida and either North Carolia or Arizona.

I think for now we have to keep as many states in play as poosible. A geographic hybrid of PA and NC would require just one more state to flip. In 2016, we got complacent because of the "Blue Wall." I don't want us to think too narrowly in 2020 because we think there is "Red Wall" behind Trump. So let's keep an eye on any pink state in that map. (Yes, even Montana. Crazier things have happened, and John Tester might be on the ticket.)

3.  If the economy stays strong, Trump will probably be re-elected. One of the most remarkable things about the Democratic gains on Tuesday is that they happened despite low-unemployment and at least some glimmer of hope that wages are beginning to finally rise. If that growth in employment lasts another two years, then Trump will be in a strong position. He will have the support of the investor class and have much more credibility than he did two years ago.

I went to bed on election night before the race was officially called for Trump. I didn't want to experience that moment live. I went to bed with Hillary having a chance for an inside straight, if some western state flipped for her. That did not happen and I leanred the awful news at about seven AM on Wednesday. This year I went to bed fearing that Trump's coalition outperformed the polls. It didn't. But it had a stronger than expected showing in Florida and Georgia.

I expect 2020 to be another nail biter. Unless there is a war or massive scandal coming out of the Mueller probe, then both bases will be fired up for their side. Lots of people will go to bed before anything has been called. I will not. I'm going to see that one through. And I encourage you to begin to think about how you want that night to turn out and what you are going to do to bring about that result. My plan is to vote first thing in the morning, then drive to Wisconsin and take little old ladies to the polls for as long as I can, whereever the Democratic candidate needs me most. If I have to live with eight years of this clown show, I'm going to do so with a clear conscience as to my own efforts.

I will close with one last snipped from my predictions post. Because it needs to be drilled into everyone's head for the next 721 days:  I will never, ever trust Florida in an election

The Fun Stuff: General Election Matchup Speculation

Things Change.

Note: a draft of this post was accidentally posted on October 5th. The final version was written on October 18, 2015.

The above map looks very different than our current political alignment. But it wasn't so long ago. That election was held 2 days before my 3rd birthday.  In 2012, only 20 states voted the same way they did that night. And only 8 states (plus DC) have given all of their electoral votes to the same party in every election of my lifetime.  (Minnesota and DC for the Dems, Alaska, Utah, Idaho, Wyoming, North Dakota, South Dakota, Kansas and Oklahoma for the GOP. Nebraska has been won by the Republican candidate in all those elections, but Obama did win one electoral vote from there in 2008.)

In my last post I made the case for why I believe the Democratic nominee is likely to win the next presidential election. My premise is that Democrats have a small but significant and growing advantage with the current presidential electorate.  To demonstrate this advantage, I categorize all of the states based on how they have voted in the past four presidential cycles.

True Blue (18 states that voted Dem in all four Bush/Obama Elections):  241
Baby Blues (3 states that voted Dem 3 out of 4 times)                                 15
Purple States (5 states that voted for the winner 4 times:                              75
Pink States (2* states that voted for the GOP 3 out of 4 times)                     27
Deep Red (23 states that voted GOP in all 4 Bush/Obama elections:         179
Note: Nebraska's second district is counted with the pink states.

So there are 256 votes likely to vote for the Democrat and 206 likely to vote for the Republican in a close election.  These 206 votes are the exact votes that Romney won in 2012.

But it's not all good news for Team Blue. In 2012 the Democrat won every close state except North Carolina. That is actually good news for the next Republican nominee. He or she will only have to really defend one state in a close election. Everywhere else they will be on offense.  So the biggest question is what is the quickest path for them to get to 270.  The simplest way to compute this is to flip the states they came closet to winning last time.

Here is every Baby Blue, Purple and Pink State, Red font indicates jurisdictions won by Mitt Romney in 2012.

Competitive States
Votes	State	Margin 2012
29	Florida	0.88%
15	N. Carolina	2.04%
18	Ohio	2.97%
13	Virgnia	3.87%
9	Colorado	5.36%
4	New Hamp.	5.58%
6	Iowa	5.81%
6	Nevada	6.68%
1	Nebraska-2nd	7.17%
5	New Mexico	10.15%
11	Indiana	10.20%

So the Republican nominee can get to 235 by just adding Florida, 253 by also adding Ohio and 266 by flipping Florida, Ohio and Virginia. Beyond that they only need to pick up one more state. Colorado, New Hampshire and Iowa were all decided by less than six points and would be competitive if the Republicans run a better candidate and campaign than they did last time.

Establishing a Baseline.

Incumbent candidates tend to be re-elected but incumbent parties running a new candidate lost more often than they win. Since 1952, Eisenhower, Nixon, Reagan, Clinton and Bush 43 were all re-elected. Jimmy Carter and Bush 41 were the only incumbents to lose. That's 5 out of 7. In that same span, Adlai Stevenson, Dick Nixon, Hubert Humphrey, Gerald Ford, Al Gore and John McCain all lost when running as candidates of the incumbent party who were not running for a second term. Only Bush 41 succeeded in that role. That's 6 out of 7.

But that is a small sample size and I think the historical trends discussed in an earlier post make the Democrats favorites to win next year.  Another factor that has to be considered is the effect of race. Some credible analysis of the election returns suggest that Barack Obama' race cost him about 4 points in the 2008 election.  (Source)  But that effect was strongest in states that were not particularly competitive, with Missouri being the only state that might have gone blue but-for the Democrats nominating a black candidate. Missouri was not competitive in 2012 and as we've already seen, Barack Obama carried most of the close states in 2012. The only additional state that probably would have been carried by Hillary Clinton or Joe Biden was North Carolina.  I also think Hillary will outperform Obama in that state next year. But for sake of simplicity, we'll stick with 332-206 as our baseline.

Match-ups with Hillary Clinton

1. Marco Rubio. I think Rubio is by far the strongest candidate that the Republicans might nominate. If he is the nominee, he has a very high likelihood of carrying Florida, which would get the Republicans to 235 electoral votes.  He would do well to consider John Kasich for a running mate and that will help in Ohio.  That would get them up to 253. Then they would have to win Virginia and one more Obama state. That is the best scenario for them and why I think Clinton is likely to choose Tim Kaine or Mark Warner as her running-mate. In any close election scenario, Virginia becomes extremely important.

Official Prediction:  Clinton 303, Rubio 253.  

2.  Jeb! Bush.  Jeb! is also more likely than not to win Florida but it will be very competitive there and I think the Bush name will hurt him in the rust belt. He also should pick Kasich to make Ohio competitive but Jeb! has already shown himself to be a lackluster (read: low-energy) candidate and I don't think he will make the race as close as handsome, charming Rubio.  I suspect he's pick up Florida but lose North Carolina.

Official Prediction: Clinton 318, Bush 220. 

3.  Donald Trump.  Let's put Trump mania in some perspective. In 2008 Dean Cain was the polling front runner for about three weeks.  Donald Trump has now led the Real Clear Politics Polling Average (RCPA) for more than 11 weeks. After the 2nd debate his poll number slipped a little. And then they rebounded.  He has a clear and often commanding lead in every national poll and almost every state-specific poll.  All very impressive.  But he averages about 25% in recent national polls. Consider that around 20 million people voted in the Republican primaries in 2008 and 2012. So 25% of the likely Republican primary voters is around 5 million people. In 2012 Mitt Romney got just over 10 million primary votes and 60 million votes in the general election.

So Trump holds sway over a plurality of Republican voters but those translate into a small sliver of the general election electorate.  If he is nominate he will get most Republican votes and some smattering of quirky independents. But his shtick would run very thin. Clinton will wipe the map with him.  She will win North Carolina, she will win Nebraska's 2nd congressional district and she will have the luxury of campaigning in places like Georgia, Missouri and Indiana in hopes of running up the score on election night. She might make Mississippi competitive if she can match the black turnout from 2012.

Official Prediction: Clinton 374, Trump 164.  (That's Clinton winning NC, NE-2, MO and GA but not Indiana. I think her husband will be an asset in MO, having carried the state twice and having been a long term governor of a neighboring state.)

4. Ted Cruz.  Senator Cruz is taking an interesting tact in this election.  He plans to run to the far-right in the primaries and hope to be the last man standing. Then he will try to increase voter turnout among working class whites in states like Pennsylvania to win the general election.  He seriously overstates his charms.  He may succeed in making some places like PA and WI competitive, but he will not flip any of them, although Ohio would probably be a true toss-up.

 (This prediction is rooted in my belief that those working class whites he thinks he speaks for are actually unlikely to be motivated to vote for a Canadian of Cuban ancestry who drips with Ivy league snobbery.)

Official Prediction: Clinton: 348, Cruz 190 (NC and NE-2 go back to blue.)

5. Carly Fiorina.  She'll fight hard and she'll sell herself well. But Hillary will savage her with commercials about outsourcing and layoffs. This will be a blow out.

Official Prediction: Clinton 363, Fiorina 175 (NC, NE-2 and GA go blue.)

6. John Kasich.

I think he'll win his home state of Ohio and be competitive in all the right places but I don't see him flipping Florida or any other battle ground. Clinton should also be able to carry NC against him.

Official Predicition: Clinton 329, Kasich 209.

And the Rest.

It's very difficult for me to imagine any of the other candidates being nominated, but I'll humor them with an official prediction.

Same Result as Trump:

7. Ben Carson: Clinton 374, Carson 164 

8. Chris Christie: Clinton 374, Christie 164 (Same result as Trump)

Same Result as Romney in 2012:

9.  Mike Huckabee: Clinton 332, Huckabee 206

Same Result as Cruz

10. Rick Santorum Clinton 348, Romney 190.  (With a lot of  wasted resources in Pennsylvania.)

11. Rand Paul Clinton 348, Romney 190

Why are the Bothering?

12.  Pataki: Clinton 385, Pataki 153 (That's the Trump result minus Indiana)

13. Jindal:  Clinton 396, Pataki 142 (Trump minus Indiana and Arizona)

14.  Lindsay Graham; Clinton 385, Pataki 153

15. Jim Gilmore: Clinton 372, Gilmore 165. (I'm giving him his home state of Virginia. This blog post is the high-watermark of Gilmore 2016.)

Match-ups With Bernie Sanders.

I think Sanders' identification as a democratic socialist is at least as big a handicap in a general election as Barack Obama's race was in the past two cycles.  If Clinton would have outperformed Obama in 2012 by carrying NC, then I think it's equally true that Sanders would have under performed by not winning Florida. So his baseline is 303 to 235.  

But I would still favor him against most likely Republican nominees. I think he cold beat Jeb! like a drum by using his family name against him and he probably salivates at the idea of running against an actual plutocrat in the form of Donald Trump.  But Rubio will play the commie card deftly. And he is a handsome young latino that listens to Jay-Z, not an aging Hippy who probably does not own an Ipod.  I think he would carry the truly close states: FL, OH, NC and VA. That gets him to 266 and he just needs to win Colorado or Iowa or Nevada.  My official hunch: he wins one of those states and sneaks into the White House.

Here are the Sanders matchups:

Rubio 272, Sanders 266

Sanders 285, Jeb! 253 (picking up FL and OH)

Sanders 332, Trump 206 (repeat of 2012 Results)

Sanders 303, Cruz 235 (Cruz wins Florida)

Sanders 285, Fiorina 253 (same result as Jeb!)

Sanders 272, Kasich 266 (Kasich wins the three big battle grounds of OH, FL and VA but fails to pick off PA or any of the western battle grounds or Iowa.)

Cason, Christie and Huckabee would probably put up a Romney and lose 302-236. Santorum and Paul might do the same although I think Bernie wold be able to pick off NC against those mediocrities. The sad-sacks would also get crushed but probably win one or two more states than they would against Hillary Clinton.

Musing on Long Shot Scenrios.

• Joe Biden's numbers would be virtually identical to Secretary Clinton's. Elizabeth Warren would start with a stronger baseline and would be most likely to win back NC and to make places like Indiana and Missour competitive. But all of them would be in a dog fight with Rubio, and a semi-comfortable place against all other potential nominees.

• If Scott Walker is the GOP running mate there will be endless talk about Wisconsin becoming competitive but I think his early exit means he won't be on anyone's short list.  (Kasich and Susan Marinez will always be a better option for any candidate on this list.)

• Trying to think of real long shot scenarios, one might be tempted to think that Hillary Clinton could win Arkansas.  She was first lady of that state for more than a decade. But that was a long time ago. In 1992 Bill Clinton won 53% of that stat's votes in a 3 way race. Four years later his share of the national vote went up by more than six points but only by one point in his home state.  Since that high water mark of 5.38% in 1996, the Democratic nominees have gotten 46, 45, 39 and 38 percent.  Obama's share of the vote dropped that much compared to Gore and Kerry. This is the most obvious case of Obama's race affecting his poll results in a non-competitive state. I think Hillary will return the Democratic number to something closer to 45 percent but 50% seems like a very long shot. Arkansas has become a pretty conventional Red State in presidential elections. I don't think the Big Dog can change that for his wife.

• Missouri is more fertile ground for an Ozark upset. In 2008 Barack Obama came within 4,000 votes of winning that state and chose not to pursue a recount because it would not have affected the election result. In 2012, he lost by 9 points. This pattern is similar to Indiana, where he won a squaker in 2008 and lost by 10 points in 2012. Nationally Obama's vote share dropped by less than 2 points, but by more than 10 points in these states. The difference of course is that Obama actively campaigned for those states in 2008. The closer polling numbers during his re-election made it necessary to spend resources elsewhere.  If Hillary is in blowout mode, she might be able to pick off Missouri, especially if she thinks she can help some marginal Democratic house candidates or the race for Roy Blount's senate seat becomes competitive. (If for instance he loses a primary to a Tea Party fringe candidate.)
• Arizona should be ripe for a competitive race. Obama lost by 7.3 points to John McCain in 2008 and by 9.0 points to Mitt Romney in 2012. That fall off, 1.7 points was slightly less than the national trend, which makes sense since he was not running against a candidate from Arizona. But Arizona politics are warped by immigration policy more than any other state. Texas is a hopelessly red state but Arizona should be competitive. Fear of illegal immigrants has led to some ghastly laws being passed and the repeated election of a clown like Joe Arapaio as sheriff responsible for a major American city. In the long run, demographics will make this a purple state and eventually a blue one, assuming the GOP doesn't change course on some of it policies. But for the immediate future, I think immigration leaves the state off of the competive list.
• New Mexico might be the decisive state in 2016. Obama carried it by double digits twice. But Bush carried the state in 2004 and lost it in 2000 by just 370 votes. If a Republican wants to win, his or her best option is this:  Romney States (206) + Ohio, Florida and Virginia (60 to get to 266. Then one more state among Colorado, Nevada, Iowa or New Mexico. New Mexico could be influenced by having a native daughter on the ticket. The GOP may need Kasich to carry Ohio but that might be all for naught if they do not win one of these small states. Susana Martinez remains a very strong candidate for the Vice-Presidency. And under at least one plausible scenario, she could win the White House for the Republicans.