Nate Silver's Blog, page 117

Election Update: Why Clinton’s Position Is Worse Than Obama’s

There’s been a potential breach of Hillary Clinton’s electoral firewall. And it’s come in New Hampshire, a state that we said a couple of weeks ago could be a good indicator of a Donald Trump comeback because of its large number of swing voters. Three new polls of New Hampshire released today showed a tied race, Trump ahead by 1 percentage point and Trump up by 5. There are some qualifications here: The poll showing Trump with a 5-point lead is from American Research Group, a pollster that’s had its issues over the years. And other recent polls of New Hampshire still show a Clinton ahead. But the race has clearly tightened in New Hampshire, with Clinton leading by only 2 to 3 percentage points in our forecast.

If Clinton lost New Hampshire but won her other firewall states, each candidate would finish with 269 electoral votes, taking the election to the House of Representatives. Or maybe not — if Clinton also lost the 2nd Congressional District of Maine, where polls show a tight race and where the demographics are unfavorable to her, Trump would win the Electoral College 270-268, probably despite losing the popular vote.

Couldn’t Clinton win Nevada to make up for the loss of New Hampshire? Or Florida? Or North Carolina? Well… of course she could. All those states remain highly competitive. The point, as we’ve said before, is just that Clinton’s so-called firewall is not very robust. If you’re only ahead in exactly enough states to win the Electoral College, and you’d lose if any one of them gets away, that’s less of a firewall and more of a rusting, chain-link fence.

To illustrate this, let’s compare Clinton’s current position in our polls-plus forecast — which gives her a 65 percent chance of winning the Electoral College — to FiveThirtyEight’s final election forecast in 2012, which gave President Obama a 91 percent chance. How could the model be so much more confident in Obama’s chances than in Clinton’s, even though we projected he’d win by 2.5 percentage points nationally and she’s ahead by 2.8? Part of it is because there are far more undecided and third-party voters this year, which could lead to a last-minute swing, or a polling error, and makes the model more cautious. But Obama’s and Clinton’s chances of winning the popular vote are relatively similar in our forecasts (76 percent for Clinton now, 86 percent for Obama then) despite that. The difference comes mostly in the Electoral College.

In the table below, I’ve run a head-to-head comparison showing how many electoral votes each candidate was projected to have at various margins of victory or defeat. For instance, Obama had a lead in states (and congressional districts) totalling 332 electoral votes in our final 2012 forecast. Clinton leads in states totalling only 272 electoral votes, just two more than the minimum she needs to win the Electoral College. This comparison is slightly unfair to Clinton because we have her down by just a fraction of a percentage point in Nevada, North Carolina and Florida — whereas Obama was ahead by just a fraction of a point in Florida — but you can see how Obama’s position was more secure. For instance, he had 303 electoral votes in states where he led by 2 percentage points or more in our final forecast, whereas Clinton has 272.

CLINTON 2016OBAMA 2012FOR STATES WHERE DEM. MARGIN IS …TOTAL ELECT. VOTESSTATES CLINTON HAS THAT OBAMA DIDN’TTOTAL ELECT. VOTESSTATES OBAMA HAD THAT CLINTON DOESN’T>= -10 points387AK SC399IN MO MT>= -9387AK SC399IN MO MT>= -8378AK GA362MT>= -7375GA359>= -6375AZ GA348>= -5359AZ348>= -4359AZ N2347>= -3341347IA>= -2323347IA OH>= -1323NC332IA OH>=_ 0272332FL IA M2 NV OH>= +1272303IA M2 NV OH>= +2272303IA M2 NV OH>= +3268CO VA281IA M2 NV NH OH>= +4239VA253M2 NV PA>= +5190247M2 MI MN PA WI>= +6185217M2 MI MN NM>= +7183216ME MI MN NM>= +8183200ME MN NM>= +9176190ME NM OR>= +10176178MEObama’s position in the Electoral College was more secure

M2 = Maine’s 2nd Congressional Distict
N2 = Nebraska’s 2nd Congressional District

Source: FiveThirtyEight’s polls-plus forecast

The differences stem from states with substantial numbers of white voters without college degrees — in particular Ohio, Iowa, Nevada and that 2nd district in Maine. (Granted, Nevada is more diverse than the other states — and it also has a history of bad polling — so I’d be careful there.) Imagine if Clinton were 3 or 4 percentage points ahead in Ohio and Iowa instead of 3 or 4 points behind, and that she were ahead by 5 or 6 points in Nevada instead of tied there. That would make for a much more robust map: losing New Hampshire would not be a problem, for instance, and she could even afford to lose Pennsylvania. Speaking of Pennsylvania, Clinton’s polling lead is narrower there than Obama’s was. It’s also much narrower in Michigan, a state that was barely even a swing state in 2012 but presents a risk to Clinton this year.

What does Clinton get in exchange? She’s polling a little better than Obama was in the highly-educated swing states of Virginia, North Carolina and Colorado. But it’s only a little bit better — not a lot better — because African-American turnout for Clinton may be lower than it was for Obama in North Carolina and Virginia, and because third-party candidates may be eating into her margin in Colorado. She also has a much better chance of winning Arizona or Georgia than Obama did. But neither Arizona or Georgia are all that close to the electoral tipping point, and Clinton will probably win them only if she’s having a strong night overall and winning all the other swing states too.

Instead, many of Clinton’s gains relative to Obama come in states that aren’t competitive at all. Clinton’s margin in our California forecast is bigger than Obama’s was four years ago, for instance. She’s polling better than him in super-well-educated Maryland and Massachusetts. And then there are her gains in red states. While Clinton’s now very unlikely to win Texas, with Trump having recovered there in recent polls, she’s still likely to make the state a lot closer than Obama did. The same goes for Utah, Alaska and South Carolina. Those could make for some extremely blue maps if Clinton were (unexpectedly at this point) to win in a landslide. But if she’s trading voters in Ohio and Iowa for those in Texas and Maryland, she’s not getting the better side of the deal.

Elections Podcast Countdown: 5 Days

 Subscribe: iTunes |ESPN App |Download |RSS |New to podcasts?

As the 2016 campaign comes to a close, the FiveThirtyEight Elections podcast team is recording daily episodes. Listen above for the latest installment, in which we discuss whether Hillary Clinton’s lead is still shrinking, one Ohio bellwether county and minimum wage ballot initiatives.

Each day, we’ll touch on some of the same basic topics:

The state of the forecastNew pollsWhat the campaign managers are emailing about todayWhat we have open in in our browser tabsHow the alternate-timeline election is lookingA listener questionOne interesting ballot initiative

For more elections coverage, check out:

The latest presidential forecastThe latest Senate forecastThe latest election updates

We’re also plugging fun videos that have nothing to do with the election:

Have questions you want answered on the podcast? Get in touch by email, on Twitter or in the comments. We will not be posting transcripts for our daily podcasts, but you can still get your daily dose of written news and analysis through Nate Silver’s election updates.

We’ll be recording daily Elections podcasts from now until Nov. 8 and posting new episodes every afternoon. Be sure to subscribe so you don’t miss a new episode.

You can listen to the episode by clicking the “play” button above or by downloading it in iTunes, through the ESPN App or on your favorite podcast platform. If you are new to podcasts, learn how to listen.

Election Update: The How-Full-Is-This-Glass Election

Hillary Clinton, like President Obama four years ago, has spent a lot of time with a chance of winning the election that is somewhere between 60 and 80 percent in our forecast. Right now, she’s squarely in the middle of that range, with a 68 percent chance according to our polls-only model and 67 percent according to polls-plus.

I’ve found that probabilities in the 70 percent range can be especially difficult to write about, because there’s the possibility of a misinterpretation in either direction. On the one hand, in a 70-30 race, you can usually cherry-pick your way to calling the race a tossup without that much effort, even when it really isn’t. We often saw that occur in 2012, when reporters and commentators consistently characterized the election as too close to call despite a fairly steady lead for Obama in swing-state polls.

On the other hand, a 70-30 race indicates that the bulk of the evidence lines up on one side of the case — in this case, that Clinton rather than Donald Trump will probably be elected president. But people can get carried away, dismiss contradictory evidence, and take a 70 percent chance to be far more of a sure thing than it really is. This year, we’ve spent a lot more time pushing back against people who seem to think Clinton has the race in the bag than we did with Obama in 2012, even though the chances of both have been similar. For most of the election, our model has seen the race as being more uncertain than in 2012, and has therefore given a better chance to the underdog (Trump) than other statistical models have. I suspect we’ve also been more bullish on Trump’s chances than the conventional wisdom has — or at least the conventional wisdom in the BosNYWash corridor — although that’s harder to quantify.

Wednesday was one of those glass-half-empty, glass-half-full days, where the same data — and there was a lot of it, with dozens of new swing-state polls — could be looked at in very different ways. On the one hand, if you thought the race was a toss-up (as some national tracking polls have it) the data convincingly argued otherwise, with Clinton holding onto leads in her “firewall” states, which are sufficient to win the Electoral College. On the other hand, Clinton’s leads are narrower than they were a few weeks ago, and any further tightening — or a modest polling error in Trump’s favor — could put her campaign in jeopardy. Furthermore, there’s a lot of disagreement in the polls, which speaks to the uncertainty in the race.

Here’s all the data — every swing-state poll we’ve added since yesterday’s Election Update — in handy chart form. It includes the latest editions of the state tracking polls from SurveyMonkey and the Alliance of Automobile Manufacturers (which are conducted by Pulse Opinion Research, an affiliate of Rasmussen Reports).

The good news for Clinton jumps out in the chart. She led in seven of eight live-caller polls — and 22 of 23 polls overall — in the states within her firewall.

But some of her leads are tenuous, especially in Pennsylvania and — although the polls wildly disagree there — Michigan. And now and then a poll will crop up showing Trump competitive or even leading in a state Clinton thought she’d put to rest: There was one such poll today in Virginia, for example. That’s not to say Trump is likely to win Pennsylvania or (especially) Virginia. But he only needs to break through in one state to broach Clinton’s firewall, whether in Pennsylvania or Michigan or New Hampshire or Colorado.

Outside her firewall, meanwhile, Clinton isn’t assured of much of anything. She got a bad set of polls in Nevada today, for example, and Florida and North Carolina are so close that Florida was split exactly 50-50 in a set of 10,000 simulations we ran at one point today. (There’s an 8 percent chance of a recount in Florida or another decisive state, incidentally.) So if Clinton were to fail in any of her firewall states, she doesn’t have particularly reassuring backup options.

Overall, the swing-state polls were about in line with what the model expected. Clinton’s win probability fell slightly on the day, in fact, although it was as a result of her continuously mediocre numbers in national tracking polls. The state polls were a wash in our model.

What’s harder to tell is whether Trump is still gaining ground, the race has stabilized, or if Clinton has even begun to rebound slightly. After several very good days of polling for Trump, today’s numbers were more equivocal. And remember that there’s a lag between a news event, when a state is polled, and when those polls hit our forecast. For example, if Clinton got especially bad numbers on Friday and Saturday after the FBI story broke, but has since recovered, we might not discover it until this weekend.

Trump’s Chance Of Victory Has Doubled In The Last Two Weeks

You’re reading Back of the Envelope, an experiment that aims to bring shorter, quicker content to FiveThirtyEight.

In this Election Update video, FiveThirtyEight Editor in Chief Nate Silver considers what impact Hillary Clinton’s latest email news is having on the polls, and he notes what data he most wants to see before Election Day.

Election Update: Yes, Donald Trump Has A Path To Victory

Tuesday was another pretty good day of polling for Donald Trump. It’s also not an easy day to characterize given the large number of polls published. You could cherry-pick and point to the poll that has Trump up 7 percentage points in North Carolina, for example, or the ABC News/Washington Post national tracking poll that has Trump up 1 point overall. And you could counter, on the Hillary Clinton side, with a poll showing her up by 11 points in Pennsylvania, or a national poll that gives her a 9-point lead.

Our model takes all this data in stride, along with all the other polls that nobody pays much attention to. And it thinks the results are most consistent with a 3- or 4-percentage point national lead for Clinton, down from a lead of about 7 points in mid-October. Trump remains an underdog, but no longer really a longshot: His Electoral College chances are 29 percent in our polls-only model — his highest probability since Oct. 2 — and 30 percent in polls-plus.

Whenever the race tightens, we get people protesting that the popular vote doesn’t matter because it’s all about the Electoral College, and that Trump has no path to 270 electoral votes. But this presumes that the states behave independently from national trends, when in fact they tend to move in tandem. We had a good illustration of this in mid-September, when in the midst of a tight race overall, about half of swing state polls showed Clinton trailing Trump, including several polls in Colorado, which would have broken Clinton’s firewall.

This time around, we haven’t seen too many of those polls in Clinton’s firewall states, such as Colorado, Pennsylvania, Wisconsin and Michigan. But that’s misleading, because we haven’t seen many high-quality polls from those states, period! We have seen lots of polls from North Carolina and Florida — for some reason, they get polled far more than any other states — and plenty of them have shown Trump gaining ground, to the point that both states are pure toss-ups right now.

So, should you expect to see polls showing Clinton behind in states like Colorado and Wisconsin? Not necessarily. Clinton probably still leads in those states, and we’d expect her to win them if she wins nationally by 4 points or so, where national polls have the race.

Here’s an illustration of that. From a set of simulations the polls-only model ran earlier this evening, I pulled the cases where Clinton won the national popular vote by 3 to 5 percentage points. In other words, we’re positing that the national polling average is about right, and seeing how the results shake out in the states:

STATEELECTORAL VOTESPROJECTED MARGINTRUMP WIN PROBABILITY (%)Clinton “firewall” (272 EV)New Mexico5-7.76Maine2-7.512Virginia13-6.33Minnesota10-5.76Wisconsin10-5.07Michigan16-4.87New Hampshire4-4.617Pennsylvania20-4.68Colorado9-4.110_Other competitive statesNevada6-1.237North Carolina15-0.446Maine CD-21-0.348Florida29-0.347Ohio18+1.266Arizona11+1.668Iowa6+1.668Nebraska CD-21+3.564Georgia16+4.992Alaska3+5.975Utah6+8.977Trump has (almost) no path if he loses the popular vote by 3-5 points

Trump’s chances are slim-to-none in this scenario. His odds are 10 percent or below in all of the Clinton firewall states except for Maine and New Hampshire — both of which our model considers more uncertain than other states for a variety of reasons. And Maine wouldn’t be enough to put Trump ahead anyway. Sure, there’s the chance that the polling in one of the other states could be wacky (maybe there’s an unexpectedly high Gary Johnson vote in Colorado, for instance). But if that happens, Clinton has some backup options in the form of Florida, North Carolina and Nevada. She’d have to get really unlucky to lose the Electoral College with a popular vote lead like the one she has now.

But the thing is, this doesn’t really have anything to do with an intrinsic advantage for Clinton in the Electoral College, or Trump’s lack of a path to 270 electoral votes. It’s just saying that if the polls are about right overall — even if they’re off in some individual states — Clinton will win. We agree with that, and that’s why Clinton’s a favorite in our model overall. The polls have her ahead.

The question is how robust Clinton’s lead would be to a modest error in the polling, or a further tightening of the race. So here’s a second set of simulations, drawn from cases in which Trump or Clinton win the national popular vote by less than 2 percentage points:

STATEELECTORAL VOTESPROJECTED MARGINTRUMP WIN PROBABILITY (%)Clinton “firewall” (272 EV)New Mexico5-4.120Maine2-3.032Virginia13-2.721Minnesota10-1.831Wisconsin10-1.037Pennsylvania20-1.039Michigan16-1.038Colorado9-0.346New Hampshire4-0.348_Other competitive statesNevada6+2.673North Carolina15+3.383Florida29+3.786Maine CD-21+4.168Ohio18+5.192Arizona11+5.491Iowa6+5.691Nebraska CD-21+7.174Georgia16+8.298Alaska3+9.584Utah6+11.682Trump has many paths if the popular vote is within 2 points

This isn’t a secure map for Clinton at all. In a race where the popular vote is roughly tied nationally, Colorado and New Hampshire are toss-ups, and Clinton’s chances are only 60 to 65 percent in Wisconsin, Michigan and Pennsylvania. She has quite a gauntlet to run through to hold her firewall, and she doesn’t have a lot of good backup options. While she could still hold on to Nevada, it doesn’t have enough electoral votes to make up for the loss of Michigan or Pennsylvania. And while she could win North Carolina or Florida if polls hold where they are now, they’d verge on being lost causes if the race shifts by another few points toward Trump. In fact, Clinton would probably lose the Electoral College in the event of a very close national popular vote.

It’s true that Trump would have to make a breakthrough somewhere, by winning at least one state in Clinton’s firewall. But that’s why it’s not only reasonable but 100 percent strategically correct for Trump to be campaigning in states such as Michigan and Wisconsin. (I’ll grant that New Mexico is more of a stretch.) Sure, Trump’s behind in these states, but he has to win somewhere where he’s behind — or he’s consigning himself to four more years in Trump Tower instead of the White House. Michigan and Wisconsin are as reasonable as any other targets: Trump isn’t any further behind in them than he is in higher-profile battleground states such as Pennsylvania, and the demographics are potentially more favorable for him.

If you want to debate a campaign’s geographic planning, Hillary Clinton spending time in Arizona is a much worse decision than Trump hanging out in Michigan or Wisconsin. Sure, she could win the state — but probably only if she’s having a strong night nationally. If the results are tight next Tuesday instead, Michigan and Wisconsin are much more likely to swing the election.

2016 College Football Predictions

Elections Podcast Countdown: 7 Days

 Subscribe: iTunes |ESPN App |Download |RSS |New to podcasts?

As the 2016 campaign comes to a close, the FiveThirtyEight Elections podcast team is recording daily episodes. Listen above for the latest installment, in which we discuss what surprises this campaign could still hold, how the model makes its final predictions and California’s ballot-measure overload.

Each day, we’ll touch on some of the same basic topics:

The state of the forecastNew pollsWhat the campaign managers are emailing about todayWhat we have open in in our browser tabsHow the alternate-timeline election is lookingA listener questionOne interesting ballot initiative

For more elections coverage, check out:

The latest presidential forecastThe latest Senate forecastThe latest election updates

We’re also plugging one fun video that has nothing to do with the election. Here’s today’s:

Have questions you want answered on the podcast? Get in touch by email, on Twitter or in the comments. We will not be posting transcripts for our daily podcasts, but you can still get your daily dose of written news and analysis through Nate Silver’s election updates.

We’ll be recording daily Elections podcasts from now until Nov. 8 and posting new episodes every afternoon. Be sure to subscribe so you don’t miss a new episode.

You can listen to the episode by clicking the “play” button above or by downloading it in iTunes, through the ESPN App or on your favorite podcast platform. If you are new to podcasts, learn how to listen.

On A Scale From 1 To 10, How Much Should Democrats Panic?

In this week’s politics chat, we wonder how safe Hillary Clinton’s lead is. The transcript below has been lightly edited.

micah (Micah Cohen, politics editor): Clinton’s lead has shrunk to a hair above 4 percentage points in our polls-only model, down from about 7 points two weeks ago. So we find ourselves in an odd position where Clinton still holds a clear lead, but it’s shrinking by the day. I’ve been getting questions from Clinton supporters wondering how panicked they should be, and while we advise everyone of all political stripes to always remain calm, let’s try to answer that question today.

How safe is Clinton’s lead/how panicked should Democrats be? As tacky as it is to cite your own tweet, I’m going to do it anyway — here’s a handy scale:

.@NateSilver538 makes case for/against Dems panicking, but there are more subtle gradations to consider:https://t.co/yehtmbSPIW pic.twitter.com/83gMwXumF0

— Micah Cohen (@micahcohen) September 24, 2016

clare.malone (Clare Malone, senior political writer): I would say that Democrats should strike a 5-6 note of panic.

micah: Interesting! I’d say 4-5.

natesilver (Nate Silver, editor in chief): A 5 or 6 sounds about right. And Republicans should be at a 7 or 8.

clare.malone: Who knows how that FBI stuff is going to strike people over the next week, right? I’d also like to note for the record that our Chia Pet Model gives Clinton a slight edge:

The Chia Pet election one week out: Clinton has slight edge; both have some unfortunate "back hair" pic.twitter.com/gycyH3PU9e

— Clare Malone (@ClareMalone) November 1, 2016

harry (Harry Enten, senior political writer): I don’t even know what that scale means … but I’ll say “unease” for Democrats and “fear” for Republicans.

micah: So, Harry, that would seem to suggest you don’t think the race will continue to tighten?

harry: All I’m doing is taking the win probabilities from our models and matching them to this 1 to 10 scale.

natesilver: Yeah, I think that’s a mistake, Harry. Because that doesn’t consider the impact of the election. If there’s a 95 percent chance that the bagel shop is closed, I’m not panicked, but if there’s a 0.2 percent chance a meteor strikes me, I am.

A 25 percent or 30 percent chance of Donald Trump winning is pretty high to begin with, and considering the consequences of his winning the election from a Democrat’s POV — actually, maybe I should revise the Democratic panic upward from a 6.

harry: I don’t care for the scale. What I think should be said is that yes, Democrats should worry about a Clinton loss. But Republicans should be more worried because Trump will probably lose.

clare.malone: The closeness of this race — i.e., many people acting as they might if there were a generic Republican nominee — is, I think, a reason for Democrats to be worried.

natesilver: If I can be a bit pedantic, the race isn’t that close. But it’s highly uncertain.

clare.malone: Pedant away.

natesilver: It’s uncertain, in part, because of the risk of a popular vote-Electoral College split. And, in part, because there are various reasons to think polling error could be high this year, such as the number of undecided voters.

You can see those forces at play in the recent tightening. Clinton hasn’t really declined very much in these latest polls. But she was at only 46 percent in national polls, and that left a little bit of wiggle room for Trump.

If we start to see Clinton creep up to 47 percent or 48 percent, Democrats can begin to panic less.

harry: Can we get into this a little here, Nathaniel? I know we’ve discussed this a bit before. Clinton is holding onto fairly strong leads in Michigan and Pennsylvania, and I’m not seeing a “path” for Trump despite some of the national polls looking really close. Yet our model has Clinton with a MUCH higher chance of winning the popular vote and losing the Electoral College than vice versa. What the heck is going on?

natesilver: Well, if she wins the national popular vote by 4 percentage points (her current margin), she wins the Electoral College almost for sure. The problem is that her path is fragile. Subtract 2-3 points from her margin, and a lot of states get awfully close. If she underperforms the polls because Trump gets a big white working-class turnout and Clinton gets a weak black turnout, then Michigan becomes competitive, for instance.

We also don’t have a lot of good polls lately in Michigan, Colorado or Wisconsin. If there are a bunch of ’em showing an 8-point lead for Clinton or whatever, obviously she’ll gain in our forecast.

harry: So what you’re saying is that there is uncertainty, but the race isn’t THAT close.

natesilver: I’m saying the polls only have to be 2-3 points off to make the Electoral College hairy for Clinton.

harry: But not harry for Clinton.

natesilver: Thanks — you’ll be here all night.

clare.malone: So we’ve been talking a lot about how low black turnout is a worry for the Clinton campaign, or Trump getting more non-college-educated whites than expected. But are there other groups that are making her path “fragile,” as you say? Like, any other groups they’re worried about weak turnout-wise?

micah: I’d maybe worry if I were Clinton that college-educated whites won’t be quite as pro-Dem as the polls suggest? Just in the sense that that’s a break with history.

clare.malone: Yeah. That seems right.

natesilver: Ohhh, I think that’s the last thing Clinton has to worry about.

clare.malone: Why? You wouldn’t worry at all?

natesilver: She’s going to get the educated white vote in big numbers. And she’s going to do well with Hispanics. The question is about African-Americans, young voters and how much turnout Trump can get among the white working class.

clare.malone: But couldn’t more conservative college-educated whites who had made peace with Clinton be turned off by the latest iteration of emails and say, “Hey, I’m not all that motivated to get out there”?

natesilver: They’re clearly coming out in big numbers in early voting. And, generally, highly educated voters are going to be more locked into their choices at an earlier point in the campaign.

clare.malone: Ok. Fair ’nuff.

micah: Most of that early voting happened before the latest FBI news.

natesilver: Contrarian-ly enough, I think the FBI news is somewhat overrated as a source of Democratic panic even though I think Trump’s chances are maybe a little bit underrated by the conventional wisdom overall.

micah: I guess, really, the question for Clinton, though, is: Will the race continue to tighten?

harry: Well, it was tightening before the FBI news.

micah: Right.

clare.malone: I want to ban the word “tighten” after the election. It’s so irritating, and it’s all you hear lately. Can’t we say, “the race is closer”? Or “the gap is closing”?

micah: What have you got against “tighten”?

clare.malone: Bad connotations. Tight pants. Tight squeeze. Getting too tight on a Friday night.

harry: The biggest worry I think for Clinton is whether she’s losing voters. It seems that she is, even if most of Trump’s gains have come from Libertarian candidate Gary Johnson and undecided voters.

micah: OK, let’s jump back to that national/state discussion: If you are Clinton and you are worried, is that worry national or state-specific?

harry: For me, it’s national. Her chance of losing if the national popular vote is where it is right now is quite small. It’s only when the national vote becomes tighter that a Clinton loss becomes more likely.

clare.malone: Well, demographically, I guess your worry nationally is black voters, as discussed above.

natesilver: There’s a lot of uncertainty at the state level, especially in states that haven’t gotten much polling. You know what the three most recent polls in Colorado say? Clinton +1, Clinton +3, Clinton +4.

Now, those aren’t great polls. But people are acting like Clinton has leads of 6-8 points everywhere in her firewall. She doesn’t, and there are really only one or two high-quality polls in a lot of these states and basically none in others.

harry: Clinton’s going back into Colorado with ad spending.

natesilver: She should have been spending money in Colorado all along.

harry: Well, I concur.

clare.malone: Can I ask a question about Trump’s spending in New Mexico? What’s the theory there? They sent out a big release today.

harry: It makes perfect sense to me, actually.

micah: Trump’s blue state travels were going to be our chat topic for today, but Nate wanted to do the Election Update on it.

He’s selfish.

clare.malone: EXPLAIN, PLEASE

harry: Here’s why: If Trump has been gaining Johnson voters from Trump, why not go after the state where Johnson is polling the best?

natesilver: I think New Mexico is a stretch, because Clinton really does seem to be doing well with Hispanic voters. But there’s a lot of uncertainty there, and it’s a cheap state to invest in.

Basically people on Twitter are like “HAHA LOSER” whenever Trump campaigns in a state that’s within Clinton’s firewall. But Trump has to win a firewall state or two in order to win the election!

harry: Yes, although the Hispanic voters in New Mexico are somewhat different than in other parts of the nation because they’ve been in the country far longer.

micah: But is he choosing from among Clinton’s firewall states wisely?

natesilver: Sure. Wisconsin and ESPECIALLY Michigan are VASTLY underrated targets for Trump.

micah: And Minnesota, right?

natesilver: Michigan is basically Pennsylvania, but with worse polling (hence more uncertainty) and probably slightly better demographics for Trump.

Minnesota’s more a case where there hasn’t been much polling. So I really don’t know about that one.

clare.malone: Yeah, I thought this map we did was really interesting regarding the possibility that Trump will overperform past GOP nominees in the Northeast and around the Great Lakes.

harry: Speaking of Michigan, a lot of Democrats in my Twitter feed are worried about that state because the primary polling wasn’t good on the Democratic side. Any thoughts on that?

clare.malone: Pure Michigan.

micah: Yeah, how ’bout that: Should Clinton supporters be more worried about the race tightening, the polls being wrong in Trump’s favor or both? (I’d vote the polls being wrong as the chief worry.)

natesilver: It’s the combo, really. With the tightening, we’re NOT at the point where “Clinton can only lose because of a catastrophic polling error.” She could lose given a relatively minor polling error if the race tightens much further.

micah: Wait, we’re already there? Or only after a bit more tightening?

natesilver: Another point or so and we’re within the “margin of Brexit.”

harry: You want to be within 3 points in the average of the final national polls if you want to win because of a polling error that is not out of this world.

natesilver: But … I don’t know that people should necessarily assume that the polls will keep tightening. The way that our model works, I’d guess that it’s more likely than not that Trump will gain a bit further as backlogged polls from the weekend come in. But there’s a world in which the polling over the weekend was a low point for Clinton as the FBI story fades from the news. You already see Google search terms for Clinton returning to normal levels, for instance.

harry: Al Gore was down 3 percentage points to George W. Bush in the final national polls in 2000, for example.

clare.malone: The metrics of this race are so hilarious to me sometimes.

harry: Can I bring up another metric so Clare can laugh herself silly?

micah: Yard signs?

harry: I’ve been following the Gallup favorable ratings. They have large sample sizes. They aren’t about “enthusiasm” (i.e., there’s no likely voter screen), and they’ve tended to track with the horse-race polls in this race. Trump hasn’t really moved up on Clinton. I just think it’s interesting.

natesilver: Closing thought: While the FBI talk has died down, Clinton hasn’t really succeeded in making Trump the center of conversation over the final week of the campaign. And Trump’s done a good job of laying low. But a week is a long time, and there may even be time for one more “momentum” shift.

The Odds Of An Electoral College-Popular Vote Split Are Increasing

We’ve written about this before, but I wanted to call your attention to it again because the possibility of an Electoral College-popular vote split keeps widening in our forecast. While there’s an outside chance that such a split could benefit Clinton if she wins the exact set of states that form her “firewall,” it’s far more likely to benefit Donald Trump, according to our forecast. Thus, as of early Monday evening, our polls-only model gave Hillary Clinton an 85 percent chance of winning the popular vote but just a 75 percent chance of winning the Electoral College. There’s roughly a 10 percent chance of Trump’s winning the White House while losing the popular vote, in other words.

As an illustration of this, we can compare Clinton’s current margins in our polls-only forecast against President Obama’s performance in 2012. Clinton — despite Trump’s recent improvement in the polls — leads by 4.7 percentage points in the national popular vote, a wider margin than Obama’s 3.9-point victory over Mitt Romney in 2012.

But Clinton is performing worse than Obama in 10 of the 12 states that were generally considered swing states in 2012. In some cases, such as Florida and Pennsylvania, the difference is negligible. She’s underperforming Obama substantially, however, in Iowa, Michigan, Ohio and Nevada and to a somewhat lesser extent in Wisconsin and Minnesota. She’s considerably outperforming Obama in Virginia and North Carolina, conversely, but that’s not enough to make up for her losses elsewhere.

OBAMA MARGIN, 2012CLINTON POLLS-ONLY MARGIN, 2016National popular vote+3.9+4.7Traditional swing statesColorado+5.4+4.6Florida+0.9+0.6Iowa+5.8-1.1Michigan+9.5+5.3Minnesota+7.7+6.1Nevada+6.7+1.4New Hampshire+5.6+5.1North Carolina-2.0+1.3Ohio+3.0-0.8Pennsylvania+5.4+5.0Virginia+3.9+7.2Wisconsin+6.9+5.5New swing statesArizona-9.0-1.0Georgia-7.8-4.4Maine+15.3+8.2Utah-47.9-8.5Clinton is underperforming Obama in traditional swing states

So how is Clinton doing better in the popular vote overall, despite failing to match Obama’s performance in most of these swing states? A lot of it is her strong performance in red states, or at least red states where a significant number of Romney voters were whites with college degrees. Thus, Clinton is putting states such as Arizona into play and — although she’s unlikely to win them — states such as Texas, Georgia and even Utah are liable to be much closer than we’re used to. Texas, in particular, can cause a potential Electoral College-popular vote skew because of its large and growing population. If the Democrat goes from losing Texas by 15 percentage points to losing it by 5 points instead, that produces a net gain of about 0.6 or 0.7 percentage points of the popular vote — larger than the margin by which Al Gore beat George W. Bush in the popular vote in 2000 — without changing the tally in the Electoral College.

Clinton’s also making gains among Hispanic voters, but she may not replicate Obama’s turnout among African-Americans. African-Americans, though, are more likely to be located in swing states than Hispanics are. Roughly half the U.S. Hispanic population is in Texas or California, another state where Clinton is likely to outperform Obama (she currently leads in California by 25 percentage points, while Obama carried the state by 23 points) without getting any additional Electoral College benefit from it.

There’s one big qualification: Our model doesn’t account for any sort of ground game advantage for Clinton in the swing states, other than to the extent that advantage is reflected in the polls. That could make a split a bit less likely than our model infers. Still, Trump’s coalition of white voters without college degrees are overrepresented in swing states, especially in the Midwest, while Clinton’s voters are not.

Elections Podcast Countdown: 8 Days

 Subscribe: iTunes |ESPN App |Download |RSS |New to podcasts?

As the 2016 campaign comes to a close, the FiveThirtyEight Elections podcast team is recording daily episodes. Listen above for the latest installment, in which we discuss how the media responded to FBI Director James Comey’s letter to Congress, which states are really in play and Colorado’s attempt to establish universal health care.

For more elections coverage, check out:

The latest presidential forecastThe latest Senate forecastThe latest election updates

We’re also plugging one fun video that has nothing to do with the election. Here’s today’s:

Have questions you want answered on the podcast? Get in touch by email, on Twitter or in the comments. We will not be posting transcripts for our daily podcasts, but you can still get your daily dose of written news and analysis through Nate Silver’s election updates.

We’ll be recording daily Elections podcasts from now until Nov. 8 and posting new episodes every afternoon. Be sure to subscribe so you don’t miss a new episode.

You can listen to the episode by clicking the “play” button above or by downloading it in iTunes, through the ESPN App or on your favorite podcast platform. If you are new to podcasts, learn how to listen.