The plane has a 20% chance of crashing by Matt Yglesias

The plane has a 20% chance of crashing

Matthew Yglesias

to me

02:07

If someone told you your plane had a 20% chance of crashing, you'd be pretty nervous.    Right now, polling for House Democrats is about as good as generic ballot polling gets. In the RealClearPolitics average, they have a 9.5-point edge in the generic ballot. FiveThirtyEight's slightly more complicated average puts it at 11.3 points. Those are really big leads. Ronald Reagan beat Jimmy Carter by 9 points. Barack Obama beat John McCain by 7. But FiveThirtyEight's read of the map is that an 11.3 percent lead in the generic ballot in early September translates to an 80 percent chance that Democrats will gain a majority in the House.    Some of that is inherent uncertainty about events and how public opinion will move in the future. Some of it is uncertainty about polling — errors in the range of 1 to 3 percent points are very common, so a 4-point error wouldn't be particularly shocking.   But a lot of it is good old-fashioned disproportionality. If Democrats win the popular vote by 4 or 5 points, the GOP will probably hold the majority. At 6 or 7 points, Democrats are favored but it's not a sure thing.    80 percent is obviously a lot more than 20; Democrats are probably going to win, and one thing we'll be talking about is what's likely to change if they do. But you wouldn't board a plane with a 20 percent chance of crashing, and it's worth looking at what's behind that other door.

A Trumpier-than-ever Congress

The key to understanding what a good GOP result in November will look like is that even though they're going to lose House seats under any scenario, if they manage to cut Democrats' overall lead to 5 points and narrowly hold the House, they are almost certainly going to pick up Senate seats.    A national political win by 5 for Democrats is in line with Republicans winning Senate races in Florida, Missouri, Indiana, and North Dakota. What's more, occasionally Trump-critical senators like John McCain, Bob Corker, and Jeff Flake will be replaced by Trump loyalists.    Beyond the mechanical impacts, the psychological impact of Republicans winning when they were expected to lose will be enormous. The surprising nature of Trump's win in 2016 was a shock to the system, and if Republicans pull out a second surprise win, the narrative that he can do no wrong politically will become entrenched.   The shrunken House margin shouldn't impede GOP governance too much, while a Republican Senate majority that is both larger and Trumpier will open up a broader playing field. That plus the psychological boost of victory means we'll get Trumpism squared.    On policy, that probably means a return to last year's congressional dalliance with "welfare reform" — basically cutting food stamps, Medicaid, housing assistance, and every other kind of means-tested program. Mitch McConnell convinced House Republicans not to push this in 2018, but with a bigger majority, it'll be easier to round up the votes, and with interest rates rising, the GOP will be under pressure to find someplace to cut spending.    But beyond policy, a come-from-behind midterm win greatly increases the odds that Trump will go through with his not-so-veiled threats to fire Jeff Sessions and Robert Mueller, pardon Paul Manafort, and bring all the investigations into his campaign to an end.   And of course, Trump's vision of a Justice Department that protects his friends and punishes his enemies is more sweeping than just the four walls of the Russia investigation. If he proves the skeptics wrong by pulling out a midterm win for the GOP, the path will be clear to implement it.   Now, of course, Democrats will probably win the House, in which case "welfare reform" is dead and firing Mueller would mean risky business in terms of congressional investigations. But 80-20 odds is far from a sure thing.

TWEET OF THE DAY

Here's former Trump attorney John Dowd denying Bob Woodward's reporting that he staged a mock interview with Trump, concluded that Trump would definitely perjure himself if he did an interview with the special counsel, and then explained to Mueller's team that Trump is just too fundamentally dishonest to speak to federal investigators.  The tweet reads: Trump's former lawyer John Dowd responds: There was no so-called "practice session" or "reenactment" of a mock interview at the Special Counsel's office. Further, I did not refer to the president as a "liar" and did not say that he was likely to end up in an "orange jumpsuit."  -- tweeted by Shimon Prokupecz at 2:23 p.m. on 4 Sep 2018

POLL OF THE DAY

The first post-primary poll of the Florida governor's race has Democrat Andrew Gillum up over his Republican opponent, Ron DeSantis, by a narrow 50-47 margin.   The stakes in this race are extremely high in terms of redistricting after the 2020 census, and a strong Democratic performance in November at least might be enough to push the state legislature to agree to expand Medicaid.   The small Gillum lead is also moderately surprising in that Trump's approval ratings have held up better in Florida than in other swing states and the incumbent Republican governor, Rick Scott, is polling well in his Senate challenge to Bill Nelson.

LISTEN TO THIS

New on The Weeds podcast: What’s at stake in the midterms   We kick off the unofficial beginning of campaign season by asking: What happens if Democrats win in November, and what happens if they lose?   ▷ Listen on art19, stitcher, or Apple Podcasts.

ONE MORE THING

If you're excited for this newsletter, help us get it out into the world. Forward this email to any folks you think might be interested in keeping up with midterm news. (They can sign up here!)    And if you have any thoughts, suggestions, or concerns, you can email me at matt@vox.com.