...
...
Today is a relatively slow day in news terms, as such days are in the Trump era, so I'll take today.
For those who track such things, here is the PredictIt betting for balance of power in Congress after the midterms. Notice a couple of things. First, the numbers don't add up to a dollar. So, hey, if you want to try to take advantage of arbitrage opportunities and peel off a few pennies, go for it. Separate markets for each outcome will give you that possibility. More importantly, you can see the differences in how the House and Senate are divided. Betting favors the Democrats somewhat in the House, but the GOP in the Senate.
What makes for a partisan tide? The problem is that they are easier to evaluate in retrospect. In the post-WWII era, we have had three elections that really stand out as wave elections-- 1994, 2006 and 2010. Whether or not to count 2008... eh... that's a separate thing. It was a presidential election in which a bunch of seats changed hands along with the presidency because of a tanking economy. I keep that in a separate category.
The question is whether or not you can predict one of these things in advance. The 2010 election? Pretty much everyone with a brain saw that coming. It was going to be UG-LY for Democrats. First-term congressman Parker Griffith from Alabama switched from D to R, figuring that the oncoming red tide would sweep him away otherwise. Instead, he lost the GOP primary. Strategic Democratic retirements abounded... go read Jacobson & Kernell's Strategy and Choice in Congressional Elections for some idea of how to understand that and how it relates to partisan tides. 2010? That was clear. It was bigger than a lot of people expected, but the fact that it was a red tide? No surprises there.
2006? Kind of the same thing. The blue wave was coming, but it was bigger than expected. The fact that the Dems were going to get the House? That was easy to predict. Getting the Senate, though, required running the table. That was a hard call. They ran the table, and won the Senate.
1994? Yeah, not a lot of people saw the GOP taking control of Congress back then. In the summer of 1994, Jack Pitney and William Connelly were about to publish a book about the GOP in the House of Representatives. They hadn't had a majority since 1954. 40 years wandering the... well, swamp, not desert, but you get the point. They decided to call the book, "Congress' Permanent Minority." Just before the book went to press, they added a question-mark to the title. Lucky them, right? (Truth in advertising-- Jack Pitney was one of my undergrad advisors). Even they didn't really think that the GOP was going to take it all in 1994, though.
What do 1994, 2006 and 2010 have in common? Not much. 2006 was the year Dubya's approval ratings tanked, mainly over the Iraq War. You can draw some parallels between 1994 and 2010 over Democratic healthcare proposals 'n such, but 2006 sticks out like a sore thumb, and with an N of 3 for waves, you got a social science problem there, buddy.
Then, what about all of the not-waves? If you really want to do this thing called, "science," you can't ignore those. You know all of those times when someone says, "this is the year we WIN!" And... bupkis. Those times matter. Remember when Newt "hey, I think I'll fuck my staffer" Gingrich decided to impeach the president over a blowjob, and led the opposition party to lose seats in a midterm election because of it? Just sayin'. Every midterm election, the opposition party says to itself, "wow, the president is TERRIBLE!!! The people will RISE UP and..." Yeah. There are a lot of not-waves to consider.
So, here's what we know, basically, from looking at all elections together-- waves and not-waves and trying to figure out what predicts what. The president's popularity matters. The more popular the president is, the better his party does. The economy matters. The better the economy is, the better the president's party does. These two things are in tension right now. The economy is doing very well. The president is not popular. That's an empirical issue because we aren't used to that distinction. We aren't used to an unpopular president and a booming economy. That creates a problem in our models because 2018 is an outlier case on the basis of that alone. That's before we factor in any other craziness.
Models predicting seat swings in House elections have gotten messy trying to assess things like how "exposed" one party is if it won too many seats in the last couple of elections, which is an issue Matt Jarvis has been trying to tackle, but even before we get into that, I don't know how we handle the empirical issue of a booming economy versus an unpopular president. Does that create a problem in our models? I'll default to the models, but with less confidence because we are dealing with an outlier case.
Then, there's the problem of the Senate. What we should take from every batshit crazy story from 2006 onward (that's the year of George "Macaca" Allen) is that one fuckin' moron saying one batshit crazy thing can turn a race. Note the dual-meaning of the word, "race." Hey, Claire McCaskill is up for reelection again! How many political lives does she have? When the zombie apocalypse comes, I'm hiding behind her. She may not be facing Todd "legitimate rape" Akin this year, but never bet against Claire "nine-lives" McCaskill. Call her Catwoman. One of the cool ones, though. Not that shit Halle Berry movie. (All due respect to Halle Berry, but that movie... just, no.)
My point is that if the Senate is 51-49, as it is right now, one idiot saying one stupid thing can shift the balance of the Senate.
You want me to make a prediction about that?
No. Just... no.
So, a "blue wave?" Here's my prediction as of now. Stupid people are going to say and do stupid things. Stupidity is an error term in my models, and that error term creates a wide range of possible outcomes. We currently live in a "reality" show, and the writers are snorting some drug that Jeff Sessions is choosing to ignore because mostly rich, white people do it. Consequently, nothing makes sense, and there is no way to predict what lunacy they will concoct. There are so many plot holes in our current events that Trump doesn't know which one to grab. What's going to happen? I can predict what happens in a formulaic, by-the-numbers story. I can't predict what will happen in a story written by drug-addled "reality" show producers. Apparently, reading too much has become a hindrance to my prognosticating abilities.
President Donald Trump is about to walk into a summit with Kim Jong Un, with no preparation. I don't have a punchline for that because it's not a joke, except in the cosmic sense. When that's reality, how do I predict anything?