In this week’s politics chat, we debate Hillary Clinton’s Electoral College strategy. The transcript below has been lightly edited.
micah (Micah Cohen, politics editor): Welcome, all. For your consideration: Should Hillary Clinton, now leading Donald Trump by about 7 percentage points, concentrate on securing her “must-win” states or try to expand her map? This question is, reportedly, currently occupying the minds of Clinton campaign staffers. So which makes more sense: Try to put traditionally red states like Texas, Utah and Missouri in play? Or reinforce the firewall?
Let’s talk overall strategy and then dive into some of the states. But first, just so we’re all on the same page: Which states are in Clinton’s firewall?
harry (Harry Enten, senior political writer): Normally, we mean Colorado, Michigan, New Hampshire, Pennsylvania, Virginia and Wisconsin. If Clinton wins those, she could lose Florida, Iowa, Nevada, North Carolina and Ohio and still win the election.
natesilver (Nate Silver, editor in chief): The firewall doesn’t include Nevada, Florida and North Carolina. And you absolutely, positively want to target those three as backup options.
The question is what states you need beyond that.
micah: OK, so let’s talk in generalities to start — better to go big or play it safe?
harry: I guess the question is what’s the purpose of this campaign. If the purpose is to win, then you reinforce. If the purpose is to win big and embarrass Trump, then you go into these red states. The national tide determines most of this, so Clinton can’t win Arizona unless she’s winning big nationally (which she is currently).
natesilver: In general, I think you plan for what the swing states would look like in a tied race. In other words, what FiveThirtyEight calls the tipping-point states — those states closest to the national margin. So that means putting the extra dollar into Florida or Pennsylvania or North Carolina and not Arizona or Texas.
harry: I think I agree with Nate, but I know a lot of other folks who think this race is over and that she should go for the kill. Another consideration is expanding to other states to help down-ballot candidates.
natesilver: Obviously, I’m assuming that the goal is to maximize the chances of winning the Electoral College. Which probably should be the goal, I think?
micah: Hmmm … not sure I agree.
natesilver: Here’s the thing, though. If Clinton beats Trump by, say, 8 percentage points, is anyone really going to care if she’s won Arizona or Georgia?
harry: If you want to have a functioning government, you want to bring as many Democrats into Congress as you can. Expanding the map probably increases the chance of doing that.
micah: Yeah, I think this comes down to a question about governing.
clare.malone (Clare Malone, senior political writer): Oh, do we care about that here?
micah: Clare! Of course.
So, here’s my argument:
If she wins, Clinton is likely to enter office with one of the worst favorability ratings ever for a newly elected president (and her approval rating will likely follow). The more she can run up the score, the more Democrats she brings along and the more of a mandate she can claim, the more likely she’ll be able to enact her agenda. That’s all extra important this year — though it probably shouldn’t be — as a guard against the inevitable arguments we will hear that she “only won because she was running against Trump.”
clare.malone: Clinton’s campaign is naturally conservative, though. I think expanding the map has a lot of risks that they’re not necessarily inclined to take.
micah: Interesting … what are the risks?
clare.malone: Running yourself too thin in ad dollars, perhaps, in markets where you really need to come through? I dunno.
micah: I’m not sure how constrained their resources are, tbh.
harry: I think the Clinton folks don’t have a money problem. You could argue that by expanding the map, she’s forcing the Trump campaign to potentially spend money it doesn’t have in places it shouldn’t need to be spending it.
clare.malone: That’s fair. Maybe it’s more staff resources in the last month I’m thinking of. If you were going to really make an all-out, last-month effort in a state that’s not typically blue, wouldn’t you have to pluck some organizers from other states to head up volunteer efforts?
You can pull a lot of in-state volunteers, sure, but in the last month of a campaign, you might not be inclined to make outside hires so much as to move around internal staff resources.
This is speculation on my part. I’m not sure how they’re operating on the HR front.
natesilver: OK, but there’s sort of a bullshit sleight-of-hand that you’re pulling there, Micah. Clinton expanding her map isn’t the same thing as maximizing Democrats’ down-ballot gains.
micah: They’re correlated, no?
natesilver: They’re correlated, but you’re possibly confusing correlation for causation.
harry: I think what I’d be doing is upping the spending on those periphery states. That isn’t Texas, but it is Arizona. The states that are pinkish — where you could perhaps move a few points to the left.
natesilver: I mean, Arizona is on the verge of being a real swing state, so maybe it’s an exception. But we can talk about that later. In general, though, I think you ought to be conservative. And if down-ballot is the consideration — well, you really just ought to give a bunch of money to the down-ballot candidates, although helping them with your turnout operation could be a boost.
But it’s not like Clinton stumping for Jason Kander in Missouri would be all that helpful to him, because he’s more popular there than she is and running as sort of an anti-establishment candidate himself.
harry: I feel like Nate’s cracking to Micah’s argument here. I can see him cracking.
clare.malone: The idea of boosting Senate or even House races seems more interesting, I agree.
micah: Yeah, I think I’m winning this argument.
clare.malone: Wait. We’re talking about down-ballot races — that wasn’t exactly your argument.
micah: It was half of it. But what about the other half of my argument: the more mushy mandate thing?
harry: Seems good enough to me. Let’s not embarrass Nate anymore.
natesilver: So to return to my earlier point: What tangible difference does it make if Clinton wins nationally by 7 points and wins Arizona by a point, versus when she wins nationally by 7 points but loses Arizona by a point?
harry: People look to the map to understand how big the victory was. We have a winner-take-all system.
micah: Yeah, if the map everyone sees on Nov. 9 is covered in blue, doesn’t that make a difference?
clare.malone: I think it’s a reasonable goal for them to want to/try to win at least one unexpected state. A spot of blue in a sea of red can be a striking visual that people walk away with.
natesilver: So should they aim for states that are physically larger because they’re more impressive on the map?
clare.malone: Hah, yes.
natesilver: So Alaska then?
natesilver: Or not Alaska because it gets shrunken down?
harry: Is this a Mercator problem? I don’t know maps.
clare.malone: Just that actually solidifying somewhere like Arizona would be a storyline, or winning Georgia — that would get talked about.
micah: Look how impressive this looks:
Compared to this:
clare.malone: But that seems overboard — to go the full monty as Micah laid out in that first map and try to win allllll those wishful-thinking states.
natesilver: Why not aim for other arbitrary goals like building a path of states that go from the Pacific to the Atlantic?
micah: That would be cool.
natesilver: You can do it with only 52 electoral votes!
clare.malone: All about that visual pop.
clare.malone: Really cinch in the heartland.
micah: OK, so let’s take Clare’s more scaled-down strategy: Beyond the firewall, which state would you target?
clare.malone: Moi? I think Arizona and Georgia are the obvious ones to target this election — big minority communities in areas that are fairly concentrated that you could try to turn out. Just pour a lot of resources into GOTV in Maricopa County and Fulton County perhaps.
micah: That makes sense.
clare.malone: Plus, in Georgia’s case, the surrounding white, affluent Atlanta suburbs — Fayette County was one of the counties that we initially identified as a bellwether, which has swung much more Democratic since 2012.
natesilver: Arizona is probably better than Georgia because its vote will be less correlated with the other Southern states. In other words, if Clinton wins Georgia, she’s almost certainly already won North Carolina and/or Florida, so it’s game over.
Whereas you could sort-of-kind-of imagine a Western surge/Midwestern collapse map that looks like this:
harry: We got a McMuffin sighting!
clare.malone: We are trolling for Utah clicks so hard these days. Trawling? I dunno.
harry: I did an interview with Utah NPR. They were very nice.
clare.malone: Of course they were.
micah: OK, so if we’re ranking Clinton’s “reach” states in terms of feasibility, seems like we’re going 1. Arizona, 2. Georgia … then what?
harry: Utah has got to be in there. Indiana, South Carolina and Missouri.
clare.malone: I’ll go with Missouri. Especially if she rides Kander coattails.
natesilver: A.L.A.S.K.A. I’m not sure why Alaska is close, and the polls there are kind of crap. But the polls there have it close.
harry: From A to U, a trip through Nate Silver’s swing state America.
natesilver: Clinton’s more likely to win Alaska than Utah, per our model.
harry: I’m betting against the model here.
natesilver: You’d bet on Clinton winning Utah before Alaska? Or Trump losing Utah before Alaska? Because those aren’t the same thing.
clare.malone: That’s like an SAT word problem.
natesilver: I think our model might be underrating Evan McMullin’s chances in Utah, but probably not Clinton’s chances.
harry: OK, let’s play this game. I think the model has a lot of conflicting data coming in from Utah. And you know it’s much harder to poll a multi-candidate race. The recent polls from Utah have shown a very weird race, and “I have a feeling,” to quote the announcer who isn’t John Candy in “Rookie of the Year.”
micah: You have a feeling that what?
clare.malone: That Mitt Romney is going to win Utah out of left field. Write-in vote. Model didn’t see it coming.
harry: I have a feeling Utah is going to be a very close race. You have two polls in the past few weeks showing a tight three-way race. And when there is a three-way race, things can get WILD.
clare.malone: So much Utah talk. This chat is officially sponsored by Utah.
micah: Yeah, let’s leave Utah.
How about Texas?
harry: I’m very skeptical on Texas?
micah: Our model shows Trump up only 7 points in Kansas — closer than Utah and just a little more Trump-leaning than Texas.
harry: I think Trump will win Kansas and Texas, although the polls in Texas are getting close.
natesilver: Texas is probably a stretch this year. But there are universes where it becomes pretty important to Democrats in four, eight, 12 years.
clare.malone: Is the whole Texas aversion based on a political myth that’s been perpetuated, or is the Democratic ground operation really sparse as to inspire great doubt?
There was that group of Obama staffers a few years back that started some kind of “turn Texas blue” organization, but I feel like Wendy Davis’s failed gubernatorial bid in 2014 finished off a lot of hopes and dreams for a while on that front.
natesilver: Yeah, it’s a bit like New Jersey in terms of having been a tease for Republicans in the past.
But the combination of college-educated whites becoming more Democratic, with the growing Latino population, makes the state pretty interesting, based on the work we’ve done.
I’ll put it this way — if Texas becomes a red-leaning state instead of a solidly red state, it sort of becomes a problem in a way for Democrats. Because it has a huge number of votes, but it doesn’t help you in the Electoral College if you’re losing it by 3 points or 5 points. In that sense, it might be a good state to invest resources in to try to push it more to behaving like a true swing state.
harry: I’d actually point to Texas as a great example of how Republicans could do well even as the country continues to become more diverse. Yes, a lot of Latinos don’t vote in Texas, but look at how well Republicans have done among those who do. In the 2010 and 2014 exit polls (I know, I know, exit polls aren’t great at measuring minority groups), the Republican gubernatorial candidates got at least 38 percent of the Latino vote! That’s one of the secrets in Texas that explains why it stays red: Republicans don’t get crushed among Latinos.
clare.malone: George W. Bush always did pretty well with Latino voters in Texas, right? And he got a bigger proportion of the national Latino vote when he ran for president than any Republican had gotten previously.
harry: He did very well.
natesilver: But the move from a Bushian GOP to a Trumpian GOP could turn off a lot of those Texas Latino Republicans.
clare.malone: Oh yeah. That’s my whole analysis.
harry: It could partially explain why Texas is closer this year. (Trump’s poor numbers with college-educated whites also probably has something to do with it.)
micah: And what’s the deal with Kansas? Did Gov. Sam Brownback just ruin the GOP there?
natesilver: I think you mean … “What’s the matter with Kansas?”
clare.malone: I believe it’s “What’s the matter with Kansas.”
clare.malone: Jinx buy me a coke.
harry: I’m out. This chat has jumped the shark. (Side note: “Happy Days” went on for many seasons after that.)
natesilver: Kansas is actually a pretty well-educated state. And the local GOP is unpopular there.
harry: Yeah, and we saw how Republicans still kept all the statewide offices there in 2014.
natesilver: Yeah, and the polls sort of blew that race.
I think Clinton probably needs a 12-point win or something nationally for Kansas to be in play. There’s another concept here too, which is that some states are more uncertain than others despite not necessarily being closer.
We don’t have much polling in Kansas, so the model is less sure about it. There actually is quite a bit of decent polling in Texas, by contrast.
That’s also why, by the way, Minnesota keeps moving up in our tipping-point index. Nobody’s really been polling it for some reason.
micah: All right, so to put a bow on this: It seems like the consensus here is that Clinton should focus on her firewall states and maybe target one to two reach states? I’m the only one who thinks she should try to play everywhere?
clare.malone: You’re just a betting soul.
But, yes, three weeks out, in a wacky race, I think they can only afford to get so creative.
harry: I think the purpose is to win the damn election. Win it.
micah: The purpose of winning, though, is to enact policy. We’re three weeks out, and Clinton has a clear lead — they can afford to push the envelope. It might help on that policy front (though that’s a question we should dive into further: Do mandates matter?).
natesilver: It’s also a bit of a confidence trick. “Clinton playing in Arizona” is a good headline, especially if polls also show a close race there. So there’s some value in that. We also haven’t tapped into the question of diminishing marginal returns. Running your first ads in Texas might be more valuable than running your 1,000th ad in Pennsylvania, even if Texas is much less likely to be the tipping-point state. So there are some arguments.
Let’s keep in mind that when the race was tight in mid-September, the battleground state polls started to look really scary for Clinton. Colorado, in particular — a state they thought was in the bag — was polling as a tie in the public polls.
Maybe their private polling had them further ahead. But it should have been a wake-up call nonetheless.
harry: I hate wakeup calls. Just let me sleep.