On Trump's Gains

The polls have tightened! Maybe you've heard, people are making kind of a big deal about it. In fact they're probably overstating the extent of the tightening, but it's pretty undeniable that some tightening has occurred. And one interesting thing about it is that it's almost entirely Trump gaining ground, rather than Clinton losing ground. On October 19th, the 538 national polling average stood at Clinton 45.4%, Trump 38.8%, a lead of 6.6% for Hillary. Today Trump's all the way up to 41.0%, a gain of 2.2%! And Clinton has fallen all the way to... 45.4%. A.k.a. the exact same place. Admittedly in the interim she went up to 46.0% and came back down again, but even since Hillary's number peaked on October 26th she's only down 0.6%, while Trump is up 1.4%. The share of voters remaining undecided or going for one of the third-party options has been falling for a while now. So let's think a bit about what this means, that Trump is gaining ground while Hillary is holding her ground.

Concerning those Polls

So, um. There's no real way to sugarcoat it: the polls right now look bad. We're mostly into the general election now; Donald Trump is the presumptive Republican nominee and Hillary Clinton is the basically-also-presumptive Democratic nominee. This match-up has been polled extensively since last summer, when it (to be more precise, the Trump side of it) first became plausible. And Clinton has led in the polls, constantly. At the worst of her email scandal last August, this got down to maybe a 1-point lead, but then it got bigger and for most of 2016 it's been pretty substantial. In late March Hillary was up by more than 10 point, on average. As recently as mid-April it was still around 9%.

Now it's 2%. This is, um, scary. The only thing that's made Trump's victory in the Republican primary anything other than utterly terrifying has been the thought that, of course, he's not going to win. Right now, if you look at the polls... yikes. It doesn't look especially slam-dunk that he's not going to win. And the question is what to make of this.

Obviously, given my nature, I'm going to say that we shouldn't worry too much, not commensurate with how bad it would be if we were entering this election with only a +2% advantage anyway. And I do think that's right. Equally obvious, though, is the fact that we should be somewhat worried, and that something about this current wave of bad polls is telling us that there is maybe more of a chance that things go the wrong way than we thought.

But the key to understanding what the polls are saying right now, I think, is that they're saying a different thing from a month ago. And from two months ago, and three, and four. That means we need to ask, what's changed? And there's a clear answer: Trump has wrapped up his nomination fight, while Clinton hasn't. And the Democratic side is getting nasty. We can see this in the polls, too. Bernie's hypothetical lead over Trump grew wider at about the same time Hillary's did, earlier in 2016, suggesting that that shift was caused by something about Trump (i.e., he got less popular). But right now, Bernie's lead hasn't budged. That tells me that the current shift is nothing to do with Trump, it's to do with Hillary. Note also that since mid-April, when it was Clinton 49%, Trump 40%, Trump's gone up just over 1%, while Clinton has dropped by nearly 6%. Again, this looks like a Clinton phenomenon, not a Trump phenomenon.

But I don't recall the last month as being a particularly bad one for Clinton. No new scandals, no major missteps. The latest rumor about the email thing is that she's been fully vindicated, though a formal announcement has yet to be made. Nothing's really been going on that would suggest Hillary's become that much more toxic a candidate... in the eyes of a disinterested observer, maybe. Not in the eyes of a Bernie supporter. And there's the rub: it seems basically certain that what's going on right now is that an awful lot of Bernie supporters aren't saying they'll vote for Hillary in the general election. (Nate Silver makes the same point in a series of tweets.)

So we basically know why the polls look like they do. The thing Hillary's people keep saying about how she's fighting two campaigns right now is correct. Trump has wrapped up the Republican side, and has gotten a bit of a boost from consolidating his party. More importantly, Hillary hasn't wrapped up her side yet, and the Democrats are if anything splintering a bit as our race draws to a close. So that's the big question of the election right now: once Hillary actually wins, and Bernie drops out, do his supporters go back to saying they'll all vote for her? Can he get them to do that? If so, then we're back to a baseline of Clinton +6% or +8% or maybe even +10%, and the election looks fairly comfortable, especially since I expect the campaign to wear well for her and ill for Trump. If not, then there's potentially a lot more danger.

I would really like to think that this is just a temporary phenomenon. Something similar happened to Obama back in 2008, when McCain had wrapped up the nomination but Hillary was still hanging on, and that too passed. And, hey--Trump got a boost when he wrapped up his nomination! And Republicans have way more reason to not support him than Democrats do to not support Hillary. Waaaaay more. You see prominent conservative pundits talking about how bad Trump is and thinking of running a third-party candidate or whatever. Bernie Sanders keeps repeating how much better than the Republicans Hillary is. It would just be bizarre if this election ends up with a Republican Party unified around Trump but a Democratic Party that can't unify around Hillary Clinton, one of its leading figures for damn near three decades now. So I remain skeptical that that's what's going to happen. Probably another month from now, the polls will start looking more like they should, and the best course of action until then is not to panic.

But I can't deny that the way the whole Sanders campaign is going right now has me pretty legitimately worried. I think he's done some real damage, and I think he's going to need to work real hard to repair it when the time comes. He'd bloody well better.

A Reality Check on the Democratic Primary

It's Tuesday. Specifically, it's Tuesday, March 15th, the Ides of March, and a key turning point in both presidential primaries. But who cares about the Republicans; this post is about the Democrats. One week ago, on Tuesday, March 8th (a.k.a. International Women's Day), the soon-to-be first female President suffered a really shocking defeat in Michigan, where polls had her up by over 20 points. Bernie ended up winning, just barely, by about 1.5%. It was shocking, and made us reconsider everything we thought we knew about how the race is going. Today (well, tomorrow, really, since it's only just barely Tuesday as I'm writing this), five big states will vote, the first and last time that will happen. It's Missouri, North Carolina, Ohio, Illinois, and Florida. And we'll see whether Michigan was a harbinger of things to come, in terms of polling error, or whether things will be basically back to normal. If the latter, we're looking at big Clinton wins in Florida and North Carolina, the last Southern states to vote, a more modest Clinton win in Ohio, a close race in Missouri, and... god knows what's going on in Illinois.

But it's worth taking stock of how things are going, because... they're going really well for Hillary. You can see how well using the FiveThirtyEight delegate targets. In almost every state, Hillary's meeting or beating her targets. And while Bernie does have a few states where he's beat his targets, they're tiny little victories. He netted one delegate versus his target in Oklahoma and Maine, two apiece in Vermont (by shutting Hillary out) and Colorado, and by five in Kansas. That's his biggest win. In Michigan, his great victory only amounted to holding Hillary to her targets. Meanwhile, Hillary has seven wins bigger than Bernie's biggest, in South Carolina (+7), Georgia (+9), Alabama (+9), Mississippi (+9), Virginia (+10), Tennessee (+11), and Texas (+22). Indeed, Tennessee all by itself is enough to cancel out all five of Bernie's wins. Put it all together and she's 88 delegates ahead of her targets, which means he's 88 delegates behind his targets. He doesn't just need to win going forward. He needs to win big. He needs to win big states by big margins. He needs to beat his targets by big margins in big states. And those targets include winning New York. Hillary's from New York.

And all this, having not delivered a single win of the sort he needs.

Here's another fact about tomorrow: it marks the half-way point of the campaign. 2006 delegates will have been awarded after tomorrow, with only 1946 to come (plus a handful here or there from non-state primaries). So after tomorrow, Bernie will need to win by the same margin he's lost by to date. Let's assume he manages to hold Hillary even in delegates. And let's also say he manages to hold her even in delegates between California and New York. All of that is generous to Bernie, I think. The 538 targets would have Hillary winning 723 of the 1413 delegates at stake in those seven states, and that has Bernie narrowly winning New York. So that would leave Bernie 223 delegates behind (as he is right now), with 1224 delegates left to be awarded. He'd need to win 724 of those delegates, or nearly 60%. Take Pennsylvania and New Jersey out of the mix, the two biggest states, where a combined delegate split also looks pretty generous for Bernie, and you're talking about only 909 delegates left. He'd need to win 557 of those. Here are the states where Bernie's won more than 60% of the delegates: Kansas and Vermont. That's his home state, and one other state. (EDIT: whoops, I forgot Maine. There's also Maine.)

If he can't win big tomorrow, bigger than it looks like he should, and if he can't win big in the four biggest states that will remain after tomorrow, his task is to win everywhere else as thoroughly as he won Kansas. (And Maine!)

That really doesn't seem very plausible. And that means that, 24 hours from now, it might start looking pretty much impossible for him to win. Unless we get another Michigan.

Congratulations to Bernie Sanders on his First Good State Result!

Woohoo! He finally beat the 538 benchmarks in a state! And it'll almost certainly be two of them by night's end. His shocking 10-point victory in Oklahoma was six points ahead of his +4% benchmark, and he looks on pace to win Colorado by about six surplus points as well, with a 17-point win.

That is officially the end of the good news for Bernie Sanders tonight. Well, okay, a 73-point win in Vermont ain't bad, even if it falls short of the +83% benchmark. But his ~15-point win in Minnesota will fall short of a +21% benchmark. Meanwhile he's failing to win two states he should've won: Massachusetts, which he should've won by about 11% and will instead lose by 2% (which I think is brutal for his campaign), and Tennessee. Oh, Tennessee. 538 thinks it should've been a very competitive state. It had other ideas, joining its Southern neighbors in giving Hillary a landslide win at +37%. Taken all by itself that would imply a 39-point national lead for Hillary. Obviously that's ridiculous, but these benchmarks are just approximations and the real results would vary a little. Which is to say, the massive surplus in Tennessee for Clinton makes up for Bernie's good results in Colorado and Oklahoma.

And then we get to the rest of the South, where he was projected to lose big but actually lost bigger. In the current approximations of the final results, he's 16 points behind his benchmark in Georgia, 20 points behind in Texas, 24 points behind in Arkansas, and 28 points behind in Alabama.

Add it all together and Hillary's running about 15 points ahead of the benchmarks in the states that voted tonight. Add in the four early states and it's... still 15%. We're getting a fairly consistent view of the race, not so much state to state but when you add it all up. Sanders has run a great campaign, and his five state wins so far really are amazing given where he came from. But he's not there yet, and at this point, with the battle-lines hardening ever more, it really doesn't look like he can get there.

And even if he does, he'll need to make up for the big wins Hillary's already banked. Which looks basically impossible given the proportional delegate allocation.

Within a few weeks it's gonna start looking like Bernie shouldn't really be in the race anymore.

There's Not a Single State Where Polling (or Voting) Suggests Sanders is Leading

FiveThirtyEight just rolled out a new version of the projections I discussed in this piece. It incorporates several new demographic factors, including an urban/rural split. One interesting thing about it is that, if the race is tied, Hillary will be winning about fifteen states, out of fifty: this implies that her support is more strongly concentrated into a few big states where she has major advantages, where Sanders has more modest advantages across a larger number of smaller states. Or to put it another way, in order to win, Sanders really needs to feel like he's winning.

And... boy does this ever not look like a race where Sanders is winning. The first three states now look way better for Hillary: instead of implying national leads of 6, 7, and 2 points respectively, Iowa, New Hampshire, and Nevada now imply leads of 19, 10, and 5. Meanwhile, they also put the current 538 polling averages into their latest chart, and... whaddaya know, Sanders is currently beating his benchmarks in a whopping zero states. You've got some states, like South Carolina, Massachusetts, Oklahoma, Texas, and North Carolina, that don't look all that different from a tied race; Hillary's beating those benchmarks by single digits. But then you've got, jeez. Sanders ought to be winning Tennessee, narrowly; he's losing it by 25 points. Similarly in Michigan, where he's 20 points behind his benchmark. And then looking a little further down the calendar, you've got more states Sanders ought to be winning, or maybe even winning big, like Utah and Wisconsin. You've got big states he's got to win, like Pennsylvania and California. Hillary's leading them all: by eight, two, twenty-two, and thirteen, respectively. There just isn't a single real bright spot for Sanders in the state polling we have right now. Overall, taking a rough average of all these future states, it looks like a Clinton +13 race right now. That's very similar to the numbers implied by the first three states that have voted, which were maybe consistent with a +11 race overall.

And this, by the way, makes me skeptical of the polls we're seeing showing more like a +5 Clinton race nationally, or the one that had Sanders leading for what is I believe the first time ever. If that's so, there has to be somewhere in this nation where he's doing better than you might expect given a tied race. But those national polls that are more optimistic for Sanders, that's basically where the single most optimistic individual state data-points for him are at. And that makes it tough to view them as anything other than outliers, or perhaps pollsters whose weighting algorithms are off.

But, we shall see.

Where We're At in the Democratic Primary

A few days ago, FiveThirtyEight published a very useful piece that essentially laid out the landscape for the 2016 Democratic primary. They used some polling data, fundraising, demographics, and Facebook popularity to figure out how we can expect each state to vote if Hillary is leading by twelve points, and alternately how we would expect them to vote if she's fallen into a tie with Bernie--i.e., the first projection shifted over by twelve points, plus or minus rounding errors. And what this lets us do is convert individual state results into an estimate of how the race stands nationally, both with actual results and with polls.

So, for example, they show Nevada as a pro-Hillary state, one that should go to her by about 3 points even if the race is tied. If in fact Nevada ends up as basically a tie tomorrow (and we have very little idea what will happen there; there have only been three polls, all low-quality ones, in recent weeks), that's a fantastic result for Bernie. Even if it comes out as Hillary +6%, the most pro-Hillary of the three recent polls, that's a very good result for him. Conversely, Hillary's got something like a 25-point lead in the South Carolina polling average, which would be even better than the +23% win they project for her with a national 12-point lead. It's to be expected that different states will give different signals, because these projections are by no means perfect. (It's like extrapolating election night results on the assumption of uniform swing: it's a good approximation but never quite holds.) Indeed it kind of looks like a decent guess for Nevada is the national-tie scenario, while a decent guess for South Carolina is the Clinton lead scenario.

One interesting thing, though, is that the two states that have voted so far both say the same thing about the race. Iowa was a tie, falling almost precisely between the two projections (+7% for Hillary if she's up big, +6% for Bernie if it's a tie). Meanwhile, though Sanders won big in New Hampshire, by 22%, he didn't win as big as he should have if the race is a tie, which would've been 29%. Both states, therefore, imply a Clinton lead of roughly 6 or 7 points nationwide. Which confirms my feeling that those results were decent for Clinton, even though they looked pretty lousy. To be sure it's remarkably that Sanders is within 6 or 7 points nationally, but that's still a pace that has Hillary winning. If that pace holds, we should see Hillary winning by nearly 10 points in Nevada and by 16 or 17 in South Carolina; more likely we'll see a similar average result between those two states but highly disparate individual signals.

In any case, it's just something to keep an eye on, as a guide to interpreting each state's results.

Another Inevitable Result of the Great Inversion

So one thing that's making the rounds today is a poll showing an ever-increasing number of people saying they wouldn't like it if their child married someone of the opposite political party. This of course shows how terrible Americans are, how acrimonious our partisan divides, how tribal our politics, etc. Because you see, in 1960, 4% of Democrats and 5% of Republicans said this. In 2008 it was 20% of Democrats and 27% of Republicans. Now it's 33% of Democrats and a whopping 49% of Republicans. People are becoming that much more ideologically intolerant, I guess.

Or not. We don't know how ideologically intolerant people were in 1960. Not from this survey data at least. Because, as I detailed in my previous post, back in 1960 party identification correlated only weakly with ideology. More to the point, there was a considerable amount of ideological overlap between the two parties. A Democrat and a Republican from Massachusetts had an awful lot more in common with each other than that Democrat would have with a Democrat from South Carolina. Indeed, it's quite likely that, at least if you go back a little further, the Massachusetts Democrat would have had more in common with the South Carolina Republican than with his southern co-partisan.

So if you had the strong feeling that you didn't want your son or daughter marrying someone whose ideology you found hateful, it just wouldn't make a lot of sense therefore to decide you didn't want them marrying someone of the opposite party. You would presumably want to be discriminating among cross-partisans, just as you would need to be discriminating among co-partisans. These days, though, if someone puts an R after their name it gives me a pretty good idea of what they're all about, or at least what they're willing to tolerate. So whether you like the practice or not, it just plain makes a lot more sense to disfavor members of the opposite party, for your children or for yourself.

Polling Reveals Major Doublethink in America

George Orwell, in 1984, defined "double-think" as "[t]he power of holding two contradictory beliefs in one's mind simultaneously, and accepting both of them." I mean, he defined it in a lot more detail than that, but that's the typical canonical statement. Well, Kevin Drum brings up a recent Gallup poll that asked people whether they thought various different things would improve the quality of governance in this country. He's mostly interested in the fact that an overwhelming majority think having more businessmen in Congress would be an improvement. That's admittedly a somewhat disconcerting result (really, guys? 18% more people think having more businessmen in Congress would be an improvement than having more women? that's pathetic) but it's not the thing I find most interesting. Here's the chart summarizing the results:

Focus on the third and fifth entries. Apparently 63% of people think that we'd be better off with more people in Congress who will pursue compromise rather than sticking hard to their principles, and 56% think we'd be better off with more people in Congress who would stick hard to their principles rather than pursuing compromise. 30% and 38%, respectively, disagree. Eeeeexcept... uhhh... those numbers add up to 119%. That is to say, a minimum of 19% of the respondents in this poll said both that X was better than Y and that Y is better than X. ...yeah.

All of which is to say, this poll is yet another piece of evidence that people are just super unhappy about the state of the world right now, in general. They're not upset at anyone in particular; or rather, they're upset at everyone. So anything you present as an alternative to the status quo is going to sound good. (Note that every single one of these got a net positive response.) That's true, apparently, even if you present two completely contradictory alternatives! (Okay I suppose they're not necessarily 100% contradictory if there are some people who fall into neither of those categories; maybe what people just want is a higher concentration of extreme compromisers and extreme hardliners, with few people in the middle of that particular spectrum. But I'm not buyin' it.) I bet if they had stuck "men" in their poll as one of the things we might put more of in Congress, they'd have gotten a positive result for that, despite the fact that the "women" question is literally the inverse of that question. 63% of their respondents literally said something that is equivalent to thinking it would be a bad thing if we had more men in Congress, but I bet you that an awful lot of those people would have said the exact same opposite had they been prompted.

So don't take polls like this too seriously.

Perhaps the Founders Would Be Disappointed. So What?

Gallup apparently has a habit of asking, on one of its late-July polls, whether the respondents believe that "the signers of the Declaration of Independence would be pleased or disappointed by the way the United States turned out." They asked this for the Fourth of July in 1999, when 55% to 44% the public believed the Founding Fathers would be disappointed. Then there was a little surge of, I dunno, patriotism? Something, anyway, because when they asked it during the summer of 2001 (note that this is prior to September of 2001, so this wasn't a rally-'round-the-flag thing) the numbers had flipped: 54% to 42%. The bounce receded a bit by 2003, with a 50%/48% split in favor of "pleased."

Well, all that's changed now: just 27% think they'd be pleased, while 71% think they'd be disappointed.

I have two comments to make about these numbers, one short and one long. First, the short one: while I understand it on a psychological level, on any rational level it is insane for the numbers to have changed that much over the last decade. Right now we're in an economic slump, sure. That has astoundingly little to do with how the country has turned out when we're talking about 250 years! I suppose some of the stuff about intractable war has become more obvious in that time, and I guess if you're a right-winger you've probably gone all Clarence Thomas on the entire Obama Presidency. But, c'mon, in both of those dimensions honestly, from where we were in 1776 we were at least 90% or 95% of the way, ten years ago, to where we are now. The country just isn't that different from its 2001 version, compared to the mutual differences to the early Republic. If you think the tipping-point was in that particular decade, well, I'd be curious to hear you defend it.

Now the long one: I'm in the 71%. I think the vast majority of the Founders would be kind of livid if they were taken on a TARDIS trip from July 4th, 1776 to July 4th, 2013. Part of the reason why is that you have a lot of different changes, each with the potential to alienate a different Founding-era constituency. Any founder particularly wedded to the idea of America as a uniformly Protestant, or even just Christian, or even just religious nation would be disappointed. Any founder particularly uncomfortable with the idea of, you know, gay people--and let's be honest, that's essentially all of 'em!--would be disappointed. Thomas Jefferson's dream of an agrarian republic? Crashed and burned. And then, of course, there's the whole contingent who were committed to the supremacy and superiority of the white race, and the larger contingent who were committed to the political supremacy of the male sex. They're both in for a shock. (Of course, some of their wives and slaves might like what they would see in 2013, but none of them signed the Declaration.)

There may be other, subtler issues that would cause some to be disappointed. Yes, many of them probably wouldn't like how powerful the federal government has become, or that we have standing armies now, or that the practice of corporal punishment of children has gone mostly out of fashion. Or whatever. And psychologically, my guess is that finding one big change that a given Founder thought was a big negative would be enough to put that guy squarely in the Disappointed camp. Some might be judicious enough to say, okay, that's not how I envisioned it, but it looks like a nice place altogether, prosperous and free overall, and maybe my vision of the future didn't do so bad, not compared to all the hardline pro-slavery guys anyway. And honestly I think Alexander Hamilton would just plain love it. But most of them, I think, would find something to hate in how the future panned out, so I think it definitely true that most of them would consider themselves Disappointed.

But, guess what? They'd be wrong! All of those changes that they would be disappointed in? They were great! Racial and gender equality is awesome! Sexual liberation is awesome! Cities are awesome! Religion is not awesome! (Okay, I guess if you're particularly religious you might not be with me on that point...) And, you know what? All you conservatives who love to talk about how the Obama Administration is betraying the Founders' ideal of limited government? If Paul Ryan's budget were passed into law tomorrow, it would be a rounding error on the tally of how far the federal government's powers have been enlarged since 1789. You wouldn't be able to tell the difference. Among other things, as the Founders probably knew much better than modern "constitutional conservatives," big spending maketh not big government, and spending cuts don't fight tyranny. If you repealed the Social Security Act of 1935 and all subsequent amendments and additions thereto, you'd change the dynamic of federal power, but last I checked the Republican Party is firmly against that course of action, which means that all those conservatives are positively thrilled that we've expanded the federal government way beyond what the Founders envisioned. They just think, of course, that we went too far a couple years back. (Though of course, they're the ones happiest about something the Founders really thought was a sign of tyrannical government, namely our marvelous modern standing army.)

Really, there's only one way in which America, 2013 is substantially worse than America, 1776: our environment has spent the intervening 237 years being degraded with a vengeance, and in ways whose full devastating impact will not be fully known for another century. Other than that, all the myriad massive changes which have swept through this nation and the entire world which the people of 1776 could not have envisioned and probably would not have liked the thought of had they known they were coming have been for the better. We are a nation larger, more populous, more prosperous, more powerful, more cohesive, more egalitarian, more democratic, and, I think, freer than any reasonable extrapolation from 1776 could have foreseen. And if the Founders don't like it, they're free to spin around in their graves all they like.

(Except Alexander Hamilton, though. Seriously, that dude was awesome, and way ahead of his time. He'd feel right at home now.)

What Vegans Did To Republicans

Apparently the later national PublicPolicyPolling survey covered a bunch of issues related to food. One of those questions, apparently, was whether people liked or disliked vegans. According to a tweet from PPP:

On our food poll we found that Republicans have a negative opinion of vegans...not sure what vegans ever did to them...

Now, the answer to this could be that Republicans are associating vegans with the kind of people who commit illegal acts, mostly crimes against property, with an animal-rights-y agenda. It's fair enough to dislike the latter, I'd say, though insofar as they avoid violence against people I think it's also fair enough not to dislike them. But I don't think this is the impetus behind Republican dislike of vegans.

Rather, I think it's that they, and we vegetarians as well, make them feel bad. Very few Republicans are vegans or vegetarians. Most Republicans like to eat meat. And I think that, for many or most of them, that isn't just because they like how it tastes or whatever. They positively like the fact that they're eating the remains of a formerly living animal. It makes them feel powerful, like they have dominion over the earth. And vegans make Republicans feel bad about this. Sometimes this is through explicit criticism in conversation, or whatever, but it doesn't even need to be. Simply by adhering to a moral code that considers something these people do on a regular basis to be wrong, vegans and vegetarians by our very being imply that these meat-eating Republicans are doing something wrong. We're the worst possible manifestation of the liberal tendency to want to control everything about a person's life: we want to control what you can eat. And that's not even wrong, entirely, though I do know some vegans who'll tell you that what you eat is a personal choice (something I don't believe, on this particular issue). And we invade the natural social-conservative terrain of moralizing (though of course, if we get into the philosophical weeds, I could say that the vegan/vegetarian moralizing is consequentialist, while the typical Republican variety is deontological, so they're very different things). Basically, simply by being vegans or vegetarians we offend the Republican idea that the right kind of people, at least, can stomp around the world doing whatever they want, and making other people feel bad for doing the wrong things. So they dislike us: it's not surprising.

Odd Racial Gaps Regarding MLK's Dream

From the recent CNN/ORC poll:

"Martin Luther King gave his famous 'I Have a Dream Speech' at a civil rights march in Washington in 1963. In your view, do you think the U.S. has fulfilled the vision King outlined in that speech, or don't you think so?"

                     All adults      Whites    Non-Whites
Has fulfilled:       50%          51%          49%
Don't think so:    47%          46%          50%

If that looks like a surprisingly small racial gap, that's because it is. Shockingly so. I've scarcely ever seen a poll on a racial issue with that small a split. Admittedly the "non-white" group isn't just the "black" group, and I'd like to see the split among African-Americans per se, but I'd be surprised if it looks too radically different. (For instance, if you assume that non-white non-blacks responded the same as whites, and that blacks are 50% of non-whites, that would double the tiny gap from 6% to roughly 12%, still not too impressive.)

This isn't the really interesting thing, though. The poll also asked those who thought we have not achieved the Dream yet whether they think we can eventually do so, or not. Overall, 27% said we could get there eventually and 22% said we couldn't, along with the original 50% who said we're there yet. Among white "no" voters, however, the breakdown was 25%-23%. Among non-white "no" voters, the figures were 31%-19%. So whites had a 51%/25%/23% line overall, while non-whites were at 49%/31%/19%. That's slightly less "yes," noticeably less "not ever," and substantially more "not yet, but eventually" response among non-whites. It's an interesting combination: more long-term optimism, but also more present-tense skepticism. I suppose that makes a lot of sense, for a group that has suffered so long and has come so far, and still has so far to go to reach full equality.

How Do You Poll the Platinum Coin?

According to PublicPolicyPolling's Twitter account:

"Public opinion running about 5:1 against the Trillion Dollar Coin on the first night of our Florida poll among those with an opinion on it."

I wonder how they worded this question. No, really, I wonder. Because I think it's an intrinsically difficult thing to poll. The economics are a bit involved, the legal analysis is highly involved, it sounds so extremely silly, and the case for doing it is so extremely contextual. If you just ask people whether they would favor or oppose the minting of a $1 trillion platinum coin to fund the government, even without any explicitly biased language, of course they'll be opposed: it's ridiculous, which is why we don't do it most of the time. But suppose you gave the following introductory paragraph:

Congress limits by statute the amount of debt the federal government can issue. Economists predict that if the debt limit were hit, and the government could not borrow enough money to fund all its obligations, the consequences could be disastrous for the world economy. The federal government is currently at this limit, and for the past month the Treasury Department has been executing financial maneuvers to keep the government funded, but those maneuvers will only be successful for about another month. Observers have speculated that the Republican-controlled House of Representatives might not raise the debt limit by this deadline. One option the Administration would have should the debt limit not be raised would be to exploit a loophole in the coinage laws that would allow the Secretary of the Treasury to mint a platinum coin with a $1 trillion denomination, and use that money to fund the government until Congress raises the debt ceiling. Do you think this would be an acceptable response by the Administration were Congress to fail to raise the debt limit in the next month?

That would pretty much spell out the logic of even considering the platinum coin in the first place. Note that it's not entirely biased, either; I very deliberately decided to refer to the mint-the-coin option as "exploiting a loophole," which is what it is. The key is really that I've framed it as a response to a crisis. I bet if you gave that introductory paragraph, the people still on the phone by the end (because, let's be honest, it's pretty damn long) wouldn't be so opposed anymore.

Except, of course, this really would be a messaged result, and while those can be interesting, as they suggest how much support something could get if it got its message out really well, it's certainly quite questionable to simply ask a poll with that much messaging and present it as a straight-up result. Perhaps you could do a comparison within the same poll, i.e. ask  the bare-bones question first and then give a little description of why in the world someone might think it was a good idea to mint a $1 trillion platinum coin, and see how much support for the coin increased between the two. But it's hard, I think, to meaningfully gauge public opinion on a topic like this that is just so damn esoteric—not, of course, because people are idiots or because they couldn't understand such things, but because most people haven't taken any economics courses or any law courses or even any public policy courses, and don't spend their time reading blogosphere posts about stuff like the law and economics of the platinum coin.

The problem is, we are both committed to the idea of a government that is responsive to public opinion and to having the government make a serious attempt to solve any and all public policy problems that come before it, no matter how esoteric or tricky. At times like this, when it honestly seems like the best policy might be the utterly ridiculous one that, until you dig into the details, seems with good reason completely insane, those two commitments are in genuine tension with one another.

In any event, they say they won't mint any coins, which looks like it's probably a good idea politically. I'm not sure this means they actually won't be minting the coin, though, because you'd have to say you weren't going to mint it until the very minute when it became necessary to avoid some sort of default; it does seem somewhat unlikely, though. Either way, I really would like to know exactly what they plan to do when Republicans, faced with the Administration's admirable "we're just not negotiating over this, pass the damn thing already" stance, just say no. Actually having some kind of default happen would be, well, not good. It would be nice to think they do have some scheme in mind to avoid it, coin or no.

Chris Christie: Actually That Popular, Or Running the George Bush Play?

No, not that George Bush. His father. The point is this PublicPolicyPolling survey, which shows Christie creaming anyone the Democrats could run against him, including Newark Mayor Cory Booker. Booker trails by 14 points, while Richard Codey, Steve Sweeney, and Barbara Buono, of whom I've never heard, trail by 22, 27, and 40 points. Bruce Springsteen even trails by 36 points, despite himself being quite popular. The reason is that Christie himself is very popular, with a 67%/25% approval rating, including a 56%/32% approval rating among, uh, Democrats. So it looks like I'm stuck with my odious Governor for another four years, right? (Yeah, I'm in the 25% of disapprovers.)

That's probably correct. However, there's a real chance, I think, that Christie's current popularity is basically a bump from the aftermath of Hurricane Sandy, and that it could fade substantially over the next year. New Jersey is, after all, a blue state, and during a campaign it should be pretty hard for Christie to retain his popularity with Democrats. If, that is, a strong candidate could be found to run against him. That's why I invoke George H.W. Bush as a model: during 1991, his approval ratings were so strong that no prominent Democrat wanted to challenge him, but when diamond-in-the-rough Bill Clinton actually ran the campaign, he found Bush's strength dissipating and the incumbent highly vulnerable. I sort of think the 1992 campaign should stand as a cautionary tale to ambitious politicians, telling them not to let outlandish popularity numbers for the incumbent in the wake of some Event from which they've received substantial good press scare them away.

But, in this case, I imagine the relevant parties won't listen. With the presumptively-going-to-open-up Senate seat in 2014, it just doesn't make sense as the place to direct your efforts, if you're a Democratic prospect looking for higher office. Which means that we'll probably end up having to nominate some sacrificial lamb-type candidate, who'll run a bit of a mediocre campaign and get slaughtered. Oh well.

Lingering Effects of First Debate Almost Gone

According to tonight's FiveThirtyEight forecast, or the most recent version of it in any case, Barack Obama has an 83.7% chance of winning Tuesday's Presidential election. The last day on which he had a higher chance of victory was October 5th, at 84.9%, a very similar state of affairs. The day before that, October 4th, was Obama's highest odds of victory ever, at 87.1%. The day before that, October 3rd, was the date of the first Presidential debate in Denver, Colorado, which essentially everyone thought Obama lost. Obviously, the October 4th forecast didn't yet reflect Obama's decline, as the polls released that day didn't include any post-Denver interviews. But starting on the Fifth, Obama plummeted, dropping all the way down to a measly 61.1% chance of winning on October 12th. But the Vice Presidential Debate, on October 11th, was seen as a win for Obama's running mate, Joe Biden, and Obama won the second and third Presidential debates rather handily, and ever since the 12th of October he's been inching his way back up in the forecast. For a while this had looked like basically a point a day, though it's accelerated of late. Long story short, Obama has clawed his way back to roughly the same odds of victory that he had after Romney's debate victory had only been processed by the polls for a single day. One more good day of polling tomorrow could easily get him back to his highest victory odds yet.

Of course, that's just the "percent chance to win" forecast. Since we've gotten a lot closer to Election Day, the implicit margin of error in the forecast has gone way down, so the actual margin that's giving Obama these same odds of winning as on October 5th is much smaller than the margin he held on that date. Specifically, right now he's projected to win 305.3 electoral votes, and to win the popular vote by 2.2%, numbers which shift just slightly to 306.3 and 2.3% if you switch from the forecast to the now-cast. On October 5th, Obama was forecast with 317.7 electoral votes and a 3.9% popular-vote margin, and had a 4.7% lead that equated to an average of 327.7 electoral votes in the October 5th now-cast. So Obama's margin has been declining in more or less lock-step with what I like to think of as the cone of declining uncertainty. (Visualize the lines from any point on the electoral-vote or popular-vote forecast graphs at date X to the point that would, if it were the November 6th forecast, give the same odds of victory. If you do this for the leading and the trailing candidate from the same date, it's a cone.)

Still, the fact remains that Mitt Romney is very close to having failed to increase his odds of winning this election over the last month, including what was by far the best night of his political life.

Early Returns on the Debate Look Good

The day or so after a major news event is a really frustrating time for poll-watchers. There's a sense that the event has probably shaken up the political landscape, but because all the polling being released was conducted before that event, it doesn't reflect it yet, not even one tiny bit. The first indications you start getting are the movements of tracking polls on the second day after the event, and even this is tricky because you have to try and guess at single-day samples from the trendlines. Still, since it's the first evidence you get, it's worth giving it a go until better evidence comes along.

So, with that in mind, Barack Obama gained two points on the margin over Mitt Romney in both the registered-voter and the likely-voter models of Gallup's tracking poll today. That would be the poll that's shown wildly better results for Romney than any other for a good long while. Obama's taken a 1-point lead in the RV model, while still trailing by 3 in the LV version, but it's two points of improvement on either end. Now, this is a seven-day tracking poll, which means among other things that getting into the weeds of the one-day samples is extra treacherous, but the simple math suggests that a 2-point shift from a new one-seventh part of the sample means a net 14-point improvement yesterday over the day it replaced, last Tuesday. That's, um, big. Even with the enormous error bars on that estimate, it's a really good sign.

Even better is the fact that Obama's net approval went up by 5 points today. Since that's a three-day tracking poll, unlike the head-to-head, the simple math suggests a 15-point improvement over the day being dropped. That's basically the same number. So we have two different data points suggesting a roughly 15-point improvement for Obama over the status quo ante. If the actual shift was one-quarter that size, it's a huge deal.

In other words, today's Gallup tracking poll results are a big deal, and they're massively bullish for Obama.

About That CNN Snap Poll...

So, the main thing I'm happy to observe in the CNN snap poll about the recently-completed Presidential debate is that, though the respondents overwhelmingly said they thought Romney won the debate, the two candidates' favorability numbers didn't budge an inch between the initial round of calls before the debate and the follow-up after it. Specifically, Barack Obama went from 49% favorable, 50% unfavorable to a 49%/49% split, while Mitt Romney went from 54% favorable to 42% unfavorable before the debate to 56%-42% afterwards. So Obama gained one net point, and Romney gained two net points. Big effing deal.

But, uh, notice anything funny about those "before" numbers? They show Romney quite popular and Obama only sort of middling. That's not what any other poll has found for the general electorate these days. I'll let the pollster.com polling averages speak for themselves: Obama's got a 51.9% favorability rating in the aggregate, against just 44.0% unfavorable, while Romney's sporting just a 43.6% favorable score against 49.1% unfavorable. So in most polling heading into this debate, Obama is a net 13.4% more popularly perceived than Mitt Romney, but in the group of debate watchers in the CNN poll Romney was 13 points more popular than Obama. That's a skewed poll for you.

Now, that doesn't mean the poll is meaningless. Among people who watched the debate, most thought Romney won. But the poll is also telling us that "those who watched the debate" may well have been a Republican-leaning bunch. But that very fact is deeply informative, or should be anyway, of our analysis of the reaction to this debate. Regardless of who's more energized to vote this year, Republicans were far more energized about this debate. And this debate still doesn't seem to have changed very many people's minds about either of the candidates. So it looks to me like, while the debate was unquestionably a "win" for Romney, it may have been sort of a trivial one. And right now, he doesn't need trivial wins, he needs wins that he leverages into big swings in the polls. So far, it looks like he didn't get one. (Of course, that statement can neither be confirmed nor rebutted until about next Monday, when the first batch of post-debate polls will start coming in.)

You Can't Talk About IVR Polling Without Talking About Rasmussen

It's often a major error to analyze some shift in American politics, especially in the demographics thereof, without asking first whether it's really a nationwide phenomenon you're observing or just a really strong Southern phenomenon. For instance, I recently saw some numbers showing that essentially all of Obama's national deficit with white working-class voters comes from the South. Surprise! Southern whites, including the not-terribly-rich ones, don't like Barack Obama. Who'd've thought?

I think this study suggesting that IVR polls, commonly known as "robo-polls" where there's no live human interviewer, just prerecorded questions that one answers by pressing buttons, are "copying" live-interviewer polls to get better accuracy struggles from somewhat of a similar situation. Their evidence is that IVR polls have really lousy accuracy when they're conducted in a race that no standard-methodology pollster has yet touched, but once a race has seen a standard pollster check it out, IVR polling is just as accurate as live-interviewer thereafter. The people at PublicPolicyPolling, an intensely prolific and typically really accurate IVR polling firm, are naturally not thrilled with this. But I think the real problem with the survey is that, just as one errs by treating all of America, South and otherwise, as being one integral nation, one errs by treating all IVR pollsters alike.

Specifically, of course, I'm talking about Rasmussen Reports. Because here's the thing: we liberal polling junkies have been noticing for several years now that there's a distinctive pattern to Rasmussen's polling. Very early in a cycle, their polls look ridiculous. Specifically, they're ridiculously skewed toward the Republican candidates. Later in the cycle, if they continue to poll at all, their polls become much more sensible, and end up being quite accurate. What it looks like, given that it's pretty well-known that Rasmussen is a deeply partisan firm with a pretty substantial history of hackishness, is that early in cycles Rasmussen is engaged in polling-as-messaging, trying to create a buzz around Republican candidates that, perhaps, no one's ever given much heed to. That happened with Scott Brown, for instance. Then, since polling firms are basically only judged on the last batch of polls they release, later in the cycle Rasmussen starts using more sensible turnout models, etc., perhaps "copying" from the consensus of the polling to get good results and maintain their reputation, such as it is.

The result is that the largest chunk of robo-polling is conducted by a firm that is deliberately being inaccurate early in the cycle, when few of the traditional pollsters are yet at work, and deliberately making itself more accurate at crunch time, when most contests have already received a whole bunch o' polling. That would easily be enough to produce the kind of data that study found, I'm guessing, so I'd like to see the study run again excluding Rasmussen. I'd be pretty surprised if the result didn't change.

It's Really, "Are We Better Off Now?"

"Are you better off now than you were four years ago?" is Ronald Reagan's famous campaign line from 1980, attacking incumbent Democratic President Jimmy Carter over the state of the economy. Of course, the answer was "yes," real per capita GDP had gone up during Carter's Presidency, though it was in a period of decline right around the election and the line probably helped Reagan win that election. Sensing that Barack Obama's greatest weakness is the continued mediocrity of the economy, Mitt Romney has taken to using the same line. The assumption seems to be that if you are better off than you were four years ago, you should vote for the incumbent, but if you're not better off, you should vote the incumbent out. Now, as I've seen Matt Yglesias point out in various posts and/or Tweets, that's a curious framework through which to view partisan electoral politics. Most of the most important aspects of individual-level well-being are not even plausibly under the control or influence of the President, or even the government, really, like romantic success or the lack thereof. Even when we narrow things down to economic well-being, there are vast amounts of idiosyncrasy and serendipity involved at the individual level. Whether I get hired, fired, promoted, or what depends at least as much on my own personal circumstances, choices, and attributes as it does the general macroeconomic climate. If I happen to have been promoted at my job because I'm really good at it, or because the guy in front of me got caught in a scandal or something, does it really make sense to interpret that as evidence that I should vote for the incumbent President?

In that light, I found this post from Dave Weigel interesting. First, as he points out, people's views of whether they're better off or not don't seem to be particularly objective or independent of partisan politics, since non-whites seem to think they are better off, whereas actual economic data begs rather strongly to differ. But what interests me is the gender gap in this question. Men split pretty damn evenly on whether or not they're better off, while a huge plurality of women say they're worse off than four years ago. I'm not entirely sure what the data are on that, though I think I remember hearing that the initial burst of recession was especially bad for men but that the recovery has helped women less than it's helped men. I'd just like to point out that women will be supporting the incumbent President in this election much, much more than men. So, something isn't working the way Romney thinks it ought to; shouldn't these women, who feel like things have gotten worse for them, be his natural constituency? I mean, it's not like the Romney campaign hasn't made that very argument on numerous occasions.

I think what's going on is that the more sensible way to interact with this question, as a matter of partisan electoral politics, is to ask whether we, as a nation, are better off now than we were four years ago. And it looks like people know it. This goes beyond the gender gap; looking at the regional data, one sees that the least happy regions are the West and the Northeast, while the South is feeling pretty good about itself. But, uh, the South ain't voting for Obama. I bet that if you asked people whether they think the country is doing better overall than it was four years ago, you'd see some rather different patterns here, and that that question would have more predictive power of people's voting intentions. Of course, one can bring a boatload of skepticism to the idea that answers to that question would be independent of people's partisan allegiances, and rightfully so. One way or another, though, it looks like voters are not taking up the Romney campaign's call to vote against Obama if your life has been rough over the last four years. Good for them.

A Really Good Night

A recent PublicPolicyPolling survey for the League of Conservation Voters in Arizona showed Jeff Flake (R) leading Richard Carmona (D) just 44-43 for the state's open-seat Senate race. That's a Republican-held seat, i.e. a Democratic pick-up opportunity. That poll makes a good occasion to start thinking about, in light of the apparently-pretty-decent Democratic convention bounce, what a very good night for Democrats might look like on Election Day. By "very good night" I basically mean for things to get a little better over the next couple of months than they even quite seem now. Obviously Obama would be winning, probably with around 350 electoral votes. The more interesting thing is gaming out the Senate and trying to estimate the House.

One can get a good handle on the Senate state of play by using the nearly-month-old FiveThirtyEight subjective forecast of all the Senate races. We observe 10 "safe Democratic" race (DE, MN, NY, RI, VT, MD, CA, PA, WA, and WV), along with just 5 "safe Republican" ones (TX, UT, MS, TN, WY). Now, obviously on a good Democratic night we'll be winning the "likely Democratic" races, NJ, CT, MI, and HI, and also the "leans Democratic" ones, OH and NM. Angus King will also pick up his race in Maine, and caucus with the Democrats for another pick-up. Then there are 11 races rated between "toss-up" and "likely Republican," with the Arizona race falling into that latter-most category. Of those 11, 7 are held by Democrats right now, and 4 by Republicans. Note, however, that those are all the Republicans' pick-up opportunities, while Democrats have already picked up the Angus King seat. So Democrats need just 6 of those 11 races to hold their position, and just 3 of them to hold the Senate, including Biden and King. Even conceding Nebraska, then, we've got 10 races each of which looks perfectly winnable on its own terms. On a really good Democratic night, we might win 8 of them, and pick up a couple of seats. If we were to treat the Arizona race as Republican +1, and assume that that's representative of the national climate in general, then we might win 9 races.

Building on our Senate majority, in other words, seems distinctly possible, and that really wasn't expected earlier in this cycle. And I've got to believe that if we manage to elect more Democrats than this class elected in 2006, we'll also take back the House of Representatives. Now, I'd like to see a whole bunch of polling from these competitive races in the aftermath of the convention, to get a sense of whether all of these knife-edge battles seem to have shifted the Democrats' way. But the Arizona poll looks really good to me (and, note, it doesn't even have ridiculously favorable Presidential numbers, with Obama trailing by 9 points), and if the rest of the country is like this right now, it's starting to feel like November 6th could very easily be a really good night.

Not Just Suggestability Making Republicans Credit Romney for Killing Bin Laden

So, PublicPolicyPolling continued its long trend of asking slightly oddball questions this past weekend by asking people in Ohio whether they think Barack Obama or Mitt Romney deserves more credit for the death of Osama bin Laden. The splashy headline is that 15% of Republicans say Romney, while just 38% gave credit to Obama. Their Twitter feed informed us that they asked this question to see if Republicans will give Obama credit for anything, and the answer seems to be no. Except, well, people will say anything if you prompt them to. Maybe they just credited Romney because the poll suggested that one might plausibly credit Romney?

But no, that's not it. Because, you see, they gave results for the two other partisan groups as well. Among Democrats, it's 86% Obama, 1% Romney. Among independents, it's 64% to 1%. It looks like raw suggestibility isn't enough to generate a non-trivial number of people saying Romney killed bin Laden. And, I mean, there are plenty of independents who don't much like Obama. But they're not trying to credit Romney, they're just slightly less willing to credit Obama than Democrats. It's just partisan Republicans who are even remotely willing to say Romney should get credit here.

So, the moral of the story is to be every bit as shocked and outraged as the headline numbers suggest you should be. It's not just suggestability, it's over-the-top partisanship.