Tag Archives: 2018 elections

An hour-by-hour guide to Election Night 2018

[A general overview of the elections is in the previous post.]

Because states close their polls at different times and count the votes at different rates, Election Night always produces the illusion of a horse race. You could just go to bed early tomorrow night and find out Wednesday what happened. The information that trickles out minute-by-minute is not actually useful to you.

But lots of us love a good horse race, and many of the rest of us won’t be able to sleep well until we know how the important races come out. So I’m going to a returns-watching party, and I suspect many of you will be glued to your TV sets as well. Here’s what to look for hour by hour.

Before 7 p.m.

[All times are Eastern Standard. You can do the math to adjust for where you live.]

No polls close before 6, and no entire state closes its polls before 7. So by common agreement, none of the networks will report their exit polls or project any races before 7. The only point in turning on your TV before 6 is if you’re just too anxious to do anything else.

If you do tune in, though, you can sometimes glean a little information indirectly. The commentators have been getting exit poll results all day, and while they can’t tell you what those results say, they aren’t obligated to say anything that will make them look stupid when the results start coming out. So if they’re having a what-if conversation, like “What if young voters do (or don’t) turn out in record numbers?” chances are that will turn out to mean something. Commentators will be trying to lay down some themes that they expect the election results to fill out.

You will also hear some “party officials are worrying about X” comments. (One of the first signs things were going badly for Democrats in 2016 was when I heard a Democrats-are-worried-about-black-turnout-in-Cleveland conversation.) Officials are worried because they’ve seen something to worry about.

At 6 EST, the first results will come in from the eastern-time-zone parts of Kentucky and Indiana. Maybe you’ll find out something about the Indiana Senate race, which supposedly is leaning towards Democrat Joe Donnelly. But unless you know a lot about Indiana, those early returns won’t tell you too much, because exit polls can’t be released until the central-time-zone parts of those states close their polls at 7.

7 p.m.

Polls close in Indiana, Kentucky, Georgia, South Carolina, Vermont, and Virginia.

There’s no chance that control of Congress will be decided before California closes its polls at 11, and if there are a lot of close races it may take much longer. So in general, what you’re looking for in the early results is the unexpected: A close race that wasn’t supposed to be close, or an surprisingly easy win somewhere. Early elections are linked in some probabilistic way to later elections, so a surprise that favors one party or the other is a sign that surprises might keep favoring that party for the rest of the evening.

You’re also looking for trends in the state exit polls that might turn out to be national trends: Did young people vote? Are Hispanics turning out? Are women outvoting men?

The only really close Senate race in this bunch is in Indiana. Probably it won’t be clear for hours who won, but if it is, the winning party is off to a good start.

Virginia has two toss-up House races: Brat vs. Spanberger in VA-7 and Riggleman vs. Cockburn in VA-5. Those could be bellwethers for the country. In VA-10, Democrat Jennifer Wexton is expected to knock off incumbent Republican Barabara Comstock. That might be the first good news of the evening.

Kentucky-6, Andy Barr against Amy McGrath, is also rated a toss-up. As I explained in the previous post, Democrats can take the House (barely) without winning any toss-ups. But this would be a nice one to get. Georgia-6, where Republican Karen Handel won a close special election last year, is also a toss-up.

One of the most interesting governor races in the country is Abrams vs. Kemp in Georgia. That race has been all about race, so the election hinges on who actually votes. Big turnout, especially big turnout among blacks, favors Abrams.

7:30 p.m.

Polls close in North Carolina, Ohio, and West Virginia.

It’s interesting that a swing state like Ohio has no toss-up House races, a sign that districts are drawn badly. West Virginia’s three districts are all predicted to go for Republicans. North Carolina-9, Harris vs. McCready, is a toss-up.

In the West Virginia, Joe Manchin is expected to hang on to one of the Democrats’ most improbable Senate seats. If he doesn’t, the slim hopes of Democrats winning a Senate majority are pretty much finished.

Ohio has a competitive governor’s race, with just the slightest of edges to Democrat Richard Cordray.

8 p.m.

This is when things get serious. Polls close in Alabama, Connecticut, Delaware, D.C., Illinois, Maine, Maryland, Massachusetts, Michigan, Mississippi, Missouri, New Hampshire, New Jersey, Oklahoma, Pennsylvania, Rhode Island, and Tennessee.

Claire McCaskill’s Senate seat in Missouri is one the Democrats need if they’re going to have any chance of a Senate majority. Democratic wins in the Mississippi or Tennessee Senate races would be upsets, but Democrats need an upset somewhere. New Jersey is a solidly blue state this year, but Senator Bob Menendez has had a long series of near-misses with corruption scandals. He’s expected to win, but the race will be much closer than it should be.

Several interesting House races are in this batch. Maine-2 has been an expensive battle that seems to be leaning to Democrat Jared Golden. A court-ordered redistricting has partially ungerrymandered Pennsylvania, giving Democrats several chances to pick up Republican seats. The toss-up is PA-1, with polls giving a slight advantage to the Republican. In PA-17, Conor Lamb, who won a special election last year, is up against another incumbent, Keith Rothfus. Lamb is expected to win.

IL-6, MI-7 are toss-ups.

8:30 p.m.

Arkansas closes its polls. The four House seats and the governorship are all expected to go to Republicans. No Senate race.

9 p.m.

Polls close in Arizona, Colorado, Kansas, Louisiana, Michigan, Minnesota, Nebraska, New Mexico, South Dakota, Texas, Wisconsin, and Wyoming.

By now the shape of the evening should be coming into focus. Either there’s a Democratic rout going on in the House and the Senate is a nail-biter for Republicans, or it’s pretty clear Republicans will hang onto the Senate and the House is going to go down to the wire.

The close Senate races are in Arizona and Texas. Beto upsetting Ted Cruz is considered unlikely, but if it happens it’s the story of the night. Kansas and Wisconsin have close governors’ races.

MN-1, NEB-2, NM-2, and TX-7 are toss-up House races.

10 p.m.

Iowa, Montana, Nevada, Utah.

Montana’s and Nevada’s Senate races are ones the Democrats need to have. In Iowa-4, Congress’ closest thing to an open white nationalist, Steve King, is expected to be re-elected. But he’s gotten bad publicity late, so you never know.

11 p.m.

North Dakota, California, Hawaii, Idaho, Oregon, Washington.

If it’s a good night for Democrats and the Senate is still in play, it probably comes down to Heidi Heitkamp’s seat in North Dakota.

If Republicans are doing better than expected, control of the House will probably hinge on CA-39 and CA-48. California has a lot of mail-in ballots that only need to be postmarked by Election Day, so these races could be in doubt into next week.

1 a.m.


If Don Young’s seat in Alaska is in doubt, odds are the Democrats have already nailed down the House.

Past 1 a.m.

At this point all the votes have been cast and all the exit polls published. We just have to wait for the counting. Probably either the House or the Senate (but probably not both) will be in doubt well into the wee hours.

How the Midterm Elections Look With One Day to Go

The most important figure in tomorrow’s election is actually not on the ballot. No matter what happens, Wednesday morning Trump will still be president. This election isn’t about getting rid of Trump, it’s about controlling him. If Democrats get the majority in  one or both houses of Congress, the country will finally get some of the checks-and-balances that the Founders thought they had written into our Constitution.

Getting rid of Trump will still depend either on the 2020 elections, or on turning up evidence of impeachable offenses so compelling that more than a dozen Republican senators will be convinced.

The Senate

Republicans currently have a 51-49 majority. Tomorrow 35 seats are up for election, so on the surface you’d think it wouldn’t be that hard for Democrats to flip two seats and take control. (They need two, because Vice President Pence casts the deciding vote in a 50-50 Senate.)

However, 26 of the contested seats already belong to Democrats, so even if they hold all those, they have to flip 2 of the 9 Republican seats. It’s a tall order. (“Why did things shake out that way?” you might wonder. That’s because 2006 — when the public finally turned against the Iraq War — was a huge Democratic year, when Democrats like Claire McCaskill in Missouri and Jon Tester in Montana won seats in very red states. 2012 was Obama’s re-election year, so they managed to hold onto those seats. 2010 and 2014, by contrast, were strong years for Republicans.)

538 gives Democrats only a 1 in 6 chance of pulling this off. Here’s how it rates the individual races:

  • 18 are solid for Democrats, meaning the Democrat has at least a 95% chance to win.
  • 4 (Smith in Minnesota, Tester in Montana, Menendez in New Jersey, Manchin in West Virginia) are likely Democratic wins, with a win likelihood over 75%.
  • 3 (Nelson in Florida, Donnelly in Indiana, McCaskill in Missouri) lean Democratic, with a 60% or better win probability.
  • 2 (Sinema in Arizona, Rosen in Nevada) are toss-ups, though in each case the Democrat has a slight edge.

Already, winning all of those at the same time seems unlikely. If each race were independent of the others, for example, winning all four of the “likely” seats would only be about a 2 out of 3 bet. All three leaning seats together would be less than 1 in 3, and the two toss-ups together would be 1 in 4. (Actual combined probabilities are quite a bit higher than that, because the races are not independent rolls of the dice. As we saw with Trump’s victory in 2016, the party that wins close races in one state is more likely to win close races in another.)

Worse, all those seats together add up to just 27. In order to take control of the Senate, Democrats would need an upset in the only lean-Republican seat: North Dakota, where incumbent Democrat Heidi Heitkamp has only a 25% chance of hanging on. If she can’t pull that off, Democrats need one of the likely-Republican seats:

  • Beto O’Rourke in Texas (23%)
  • Phil Bredesen in Tennessee (20%)
  • Mike Espy in Mississippi (12%)

Unless at least one of those longshots comes in, Republicans hold control of the Senate. (Again, the races aren’t independent, which is how the odds for a Democratic majority can be as high as 1 in 6.)

The House

Democrats gaining control of the House is a much more doable job. All 435 seats in the House of Representatives are in play. Republicans currently have a 235-193 majority, with 7 seats vacant. Of the vacancies, two were formerly Democratic and five Republican. So you will sometimes hear that in order to gain a majority Democrats need to pick up 23 seats (if you count two of the vacancies as Democratic seats) or 25 seats (if you don’t).

In order to get a majority (218 out of 435), Democrats need only win the seats 538 rates as leaning their way:

  • 193 solid Democratic
  • 17 likely
  • 10 leaning

That adds up to 220. In addition, there are 18 toss-ups and 13 lean-Republican seats within range. Overall, that gives Democrats a 7 in 8 chance of winning the House. If everything breaks in their favor, they could have as large a majority as the Republicans have now.

House races, though, are unlikely to all go according to script. First, there are just so many of them that some longshot candidate is going to win somewhere. And second, House races aren’t polled as aggressively as Senate races, so some last-minute local factors could be overlooked. Somewhere, a district the media stopped paying attention to months ago is going to produce an upset.

As I’ll discuss in the hour-by-hour guide (the next post), you want to watch for toss-up or leaning seats in states where the polls close early. That will give the first indication of whether this is going to be a nail-biter or an easy Democratic win.

Governorships and state legislatures

If Democrats gain some governorships and control of some state legislatures, they’ll have a chance to undo the extreme gerrymandering that allows Republicans to maintain minority rule. (Last year, Democrats outpolled Republicans by 9% in the elections for the Virginia House of Delegates. But Republicans kept control.)

State government becomes more important as the federal government stops protecting civil rights and the environment. And if the now-more-conservative-than-ever Supreme Court overturns Roe v Wade, whether a woman can get an abortion may be up to the states.

Governors are being elected in 36 states, and at least some legislators are being elected in every state. 538 expects Republicans to win slightly more states (winding up with 26 governorships), but Democrats to wind up governing a larger percentage of the population (62%).

The closest races are in Iowa, Nevada, Ohio, Georgia, and Kansas. The two most interesting races, to me anyway, are Georgia and Florida, where Stacey Abrams and Andrew Gillum have a chance to be their states’ first black governors. (Abrams would be the first black woman governor in any state.) Abrams’ race is leaning towards her opponent, while 538 gives Gillum a 76% chance to win. Personally, I will find it very satisfying if Wisconsin finally boots out Koch puppet Scott Walker, which is the way the race is leaning.

As for state legislatures, I don’t know what to tell you, since there are very few published polls.

Ballot propositions

538 has a good rundown of the most important ones.

In general, ballot propositions cause more problems than they solve. There’s a reason we elect representatives who can focus on the issues full-time, rather settle everything by direct democracy. But in states that are heavily gerrymandered, or where running for office requires the kind of money you can only get from special interests, a ballot proposition might be the only way for the majority to make its will felt.

Michigan is one of those heavily-gerrymandered states.

Last year, Michigan Democrats won more overall votes for state House than Republicans. It was by a whisper, about half of one percentage point. But Democrats got walloped in the race that counts, as the GOP swept 63 of 110 seats.

Proposition 2 would create a non-partisan commission to draw districts for both the legislature and Michigan’s congressional districts. The commission would be given strict criteria to meet. Other anti-gerrymandering proposals are on the ballot in Colorado, Utah, and Missouri.

Given the racial biases in our justice system and the correspondingly high incarceration rate for non-whites, one way to make sure whites hold onto political power as long as possible is to keep felons from voting, even after they have served their sentences. Florida is one of the worst states for this form of voter suppression, with 10% of the voting-age population disenfranchised. Amendment 4 would give felons back their voting rights after their sentences end, except for murderers and sex offenders. (Yes, that is the proposition John Oliver was telling you about.)

Nevada, Maryland, and Michigan have propositions that would make it easier to register to vote, while Arkansas and North Carolina would make it harder to vote by requiring a photo ID.

A California proposition would repeal a number of gas taxes. A Washington proposition would create a carbon tax. Arizona’s Prop 127 would force utilities to get half their power from renewable energy by 2030.

Idaho, Nebraska, and Utah all have propositions to expand Medicaid.

Here in Massachusetts, the proposition I care most about is #3, which would protect transgender rights. I’m for it.

What Did Virginia Teach Us?

For weeks on this blog, I’d been fretting about the Virginia governor’s race. If we were really in the midst of a Democratic surge that might turn 2018 into a wave election, Virginia shouldn’t have been this tense. Hillary had won there last year (by 5%, aided by a Virginian VP) and Democrats already held the governorship (thanks to Terry McAuliffe’s 2.6% win in 2013). Maybe that didn’t point to a landslide, but surely we couldn’t lose.

Or maybe we could, or at least it looked that way for a while. Polls were averaging out to a 3% Northam advantage, with some showing Gillespie ahead. And Gillespie had been closing with a disturbingly Trumpish message: Northam was going to let Hispanic criminal gangs run wild in Virginia, while Gillespie would protect the monuments celebrating all those heroic Confederate defenders of slavery. And kneeling athletes were relevant somehow.

If he pulled off the upset, Gillespie’s campaign seemed likely to become a model for Republican candidates to keep on Trumping in 2018: Count on fear and race-baiting to bring the base out to vote, and hope Democrats stay home.

It didn’t work out that way. Northam won by 9%. That’s a big enough win to keep the 2018-wave narrative intact: Northam didn’t just repeat Hillary’s showing, he nearly doubled her margin, and more than tripled McAuliffe’s.

But beyond the horse-race aspect, what did we learn?

White identity politics isn’t enough. The clearest lesson is for Republicans: In 2016, Trumpism had two pieces, not just one. Yes, white Christian identity politics was a big chunk of it, but economic populism was another big chunk.

It wasn’t all just building a wall and banning Muslims and winking at the alt-Right. Trump was also going to bring factory and mining jobs back to America’s small towns. He had a healthcare plan that was going to “take care of everybody … the government’s going to pay for it”. He was going to protect Social Security, Medicare, and Medicaid. He would get tough with China and Mexico — not because the Chinese and Mexicans aren’t white, but because they had perpetrated “the greatest jobs theft in the history of the world“.

Trump voters who were attracted to his white Christian identity politics should still be happy with him. He put Gorsuch on the Supreme Court, he’s got ICE out terrorizing Hispanic immigrants, and he never misses a chance to stand up to ungrateful blacks, defend the Confederacy, or blame Muslims or immigrants (or best of all, Muslim immigrants) for America’s problems.

But the economic-populist side of Trumpism hasn’t been seen since the election.

Health care bills he endorsed would cut billions in Medicaid funding over the years, his tax plan is a bonanza for the wealthy, the budget the GOP passed to facilitate that tax plan cuts Medicare by billions, and Trump’s own budget proposal included billions in Social Security cuts.

Buy American, hire American” has proven to be an empty slogan, and the executive order that supposedly implements it has no real substance. (Not even Trump’s own companies live by it.) Tough on China? Not so much. In April he broke his promise to label China a currency manipulator, and when he went there this week.

Mr. Trump projected an air of deference to China that was almost unheard-of for a visiting American president. Far from attacking Mr. Xi on trade, Mr. Trump saluted him for leading a country that he said had left the United States “so far behind.” He said he could not blame the Chinese for taking advantage of weak American trade policy.

This is, in part, a consequence of his saber-rattling against North Korea: He needs China’s cooperation there, so he’ll have to give in to them on trade.

And Mexico? Let’s just say that they’re not paying for the Wall. Renegotiating NAFTA is proving to be a lot more difficult than just “getting tough” and making bigger demands. If the agreement winds up getting scrapped, the big loser will be American farmers and precisely those rural communities that were counting on Trump for help.

Given that Trump himself has abandoned economic populism, the only “Trumpism” left for Gillespie to adopt was the white Christian identity part. And while that stuff is really powerful for some Americans, those people aren’t a majority. Not in Virginia and not in America as a whole.

I’m not sure how Republicans running in 2018 can deal with this problem. What the Republican majority in Congress has been all about (with Trump’s blessing, for the most part) is traditional Republican trickle-down economics. 2016 Trumpism was unified by an I’m-going-to-protect-you theme, protecting his voters on the one hand from a future where white Christians are a minority, and on the other from a convergence of foreign competition, corporations who have no loyalty to their workers, and economic trends moving against them. A 2018 of message of “I protected you from Mexicans and Muslims and transgender people in your bathroom, but I tried to take away your health insurance and gave your boss a big tax cut” doesn’t hang together nearly as well.

Unity, calmness, confidence. On the Democratic side, the lesson is fuzzier, but I think there’s still something to learn.

The progressive/centrist split in the national party tried to project itself onto the primary, but it never really took. Superficially, the Bernie-backed progressive (Tom Perrillo) lost to the establishment candidate (Northam), but the divide between them was never that large, and Perrillo supported Northam in the general election seemingly without reservations. Northam, in turn, endorsed a number of progressive causes: $15 minimum wage, free community college, restoring voting rights to felons who have served their time, and Medicaid expansion.

Northam took advantage of his comforting image as a pediatrician, and talked calmly about jobs, healthcare, and education. He appealed to traditional Virginia gentility, saying of Trump “We’re not letting him bring his hate into Virginia.”, and (on just about every issue) talking about all Virginians working together to find solutions. That, of course, is a traditional political bromide, but it contrasts nicely with scare-mongering about immigrants.

In part, Northam did that because he had to — it’s hard to picture him shouting and rabble-rousing. But I wonder if that isn’t the right approach: progressive positions on issues, expressed calmly in terms of traditional American values like justice and fairness, rather than in a way that makes them sound radical. Northam’s manner projected confidence that our problems are solvable if we work together — and not solvable if we let people take advantage of our worst instincts and turn us against each other.

Whatever approach Democratic candidates take in 2018, it needs to take advantage of the hole in the Republican message: The vague economic populism of 2016 was a mirage. Republicans have been telling white Christians who to blame for their problems, but not offering viable solutions.

Where it shows up, and where it doesn’t. The exit polls can be sliced and diced all sorts of ways, but here’s what jumped out at me: Gillespie slightly exceeded Trump’s totals among low-income households (under $50K annually) and high-income households (over $100K), but Northam clobbered him among middle-income households. In the $50-$100K bracket, Trump edged Clinton 49%-47% in 2016, but Northam beat Gillespie 57%-41% Tuesday. That’s what turned an overall 5% Clinton margin into a 9% Northam margin.

So I went back and looked at the 2013 exit polls: The Republican candidate won the $50K-$99K households 51%-43%, almost the exact mirror image of the 2017 result. So maybe this is a trend.

I could imagine a lot of reasons: Maybe the middle-income people who changed their minds have health insurance, but aren’t sure they can keep it. They value education, but have to depend on the public schools. They plan to send their kids to college, but aren’t sure they’ll be able afford it. They’re not angry and looking for someone to blame, but they are worried and looking for a reason to hope.

I don’t have a good explanation for why the low-income households haven’t shifted. Clinton carried them 53%-41%, and Northam’s margin was about the same: 56%-43%. Maybe a more dramatic message would have helped Northam here; I don’t know.

RCP’s Ross Baird found another interesting way to look at the results: He examined Virginia’s five “pivot counties”, counties that went Obama-Obama-Trump in the last three presidential elections: They were dead heats Tuesday. What most pivot counties have in common, he says, is that “more businesses have died there than have been born … despite a net increase in entrepreneurial activity across the country since the end of the Great Recession.”

That suggests that Democrats still haven’t completed the sale. Trump may have lost his shine, but there are still Obama voters Democrats aren’t reaching.