How Iowa Voted

Julia, on your point about what share of the vote Rubio needs to get some good post-Iowa coverage, I’ll put it this way: It looks like Rubio will do well enough tonight to, at the very least, give party elites a good pretense for rallying behind him. If they fail to do so, that could be a sign that there’s something about Rubio that they don’t like, or that the party remains extremely disorganized. But the reported Tim Scott endorsement in South Carolina is a good start.

This is a night when candidates’ dreams are fulfilled or dashed … but it also gives us a chance to look at the what-ifs of the entire political process. NPR did a story on which states match America’s demographics closely — Illinois was the closest on race, Pennsylvania on income. I don’t know that the primary season would have changed if, for example, those two states were the first and second primaries in the nation. It seems to me that Donald Trump’s deft use of interview airtime — and apparent inability to be ruffled by fact-checking — had a greater influence on the arc of the GOP race so far than any other factor. Many nations that the United States considers key strategic and economic partners are parliamentary democracies, where the nation’s leader is the leader of the winning party or coalition. One of those is the United Kingdom. Less than a week ago, BBC New York correspondent Nick Bryant penned an opinion piece about our political process that says, “For international onlookers, it [the U.S. primary process] can seem freakish and bizarre: a long-running farce populated by cartoonish characters, which works as entertainment but is a poor advertisement for American democracy.” The article’s title: “Does America need to change how it elects its presidents?” What do you think of the question and what would you change, if anything?

Q: I wonder what [Jeb] Bush will do now. Ignore Iowa results? Double down and try to take Rubio down with him? Drop? — commenter Jonathan Smith A: I think that Bush won’t drop out yet — there’s too much money floating around in that campaign not to hoof it out for a little while yet, and there’s a certain amount of animosity he’s got with Rubio that adds some personal incentive. So … it’s probably closer to a double down scenario. If you navigate over to the New Hampshire Union Leader’s website, it’s plastered with Bush ads. There’s fight in him yet, so I’d say.

Pulling folded slips of paper ballots from a plastic post-office bin might seem quaint — as if we are watching a vote being counted in Colonial Williamsburg — but it’s actually archaic. The process of counting these slips is infuriatingly slow and extremely error-prone. The people who do the counting inevitably come up with different counts that have to be reconciled; piles of paper fall over (I winced every time the guy on MSNBC dropped a ballot on the floor), and the ballots are sometimes illegible. Iowa traditionalists will undoubtedly point to well-known flaws in digital voting systems, but modern versions of those systems are vastly more accurate and efficient than hand-counting. (As the Brennan Center pointed out last year, the main problem is that too many cities and states are using out-of-date electronic systems and need to upgrade to the new models.) As Hayley Munguia of FiveThirtyEight wrote on our site today, the Microsoft counting system now in use by Democrats is only as accurate as the manual head-counts used in countless school gyms and auditoriums across Iowa. But Iowa is very sentimental about its ancient caucus system, convinced that it is an authentic expression of rural, communal values, and there is no sign that it is likely to change.

So on this point about Rubio … This goes back to what Harry and Dave were saying about Rubio having endorsements waiting in the wings. One question I’ve been thinking about a lot is: What would knock this race out of stasis, especially for Republicans? How well does Rubio have to do for that signal of viability to be meaningful? I’m looking at numbers that have him at about 20 percent, with Cruz and Trump closer to 30. This is obviously getting into fuzzy terrain, but I wonder if that passes the threshold.

The good news for Marco Rubio so far is that he’s doing better than expected in far northwestern Iowa, which went super-heavily for Rick Santorum and spurned Mitt Romney in 2012. Rubio’s over 20 percent in Plymouth, Lyon and Sioux counties and is actually ​leading for the time being in O’Brien county. The bad news is that he’s not doing as well as expected in areas with high shares of degree-holders. He’s in third place in Polk (Des Moines), Story (Ames) and Scott (Davenport) so far. It’s still early and this could change, but his “Mayor of Ankeny” strategy isn’t all that apparent on the map yet.

With 19 percent of Iowa precincts reporting, Donald Trump has 27.1 percent of the Iowa vote. That’s not a bad result by any means: Trump trails Ted Cruz by just 3 points and could very easily win the state. Still, a case can be made that (contra the pundit conventional wisdom at the time) Trump was mistaken to have skipped last week’s debate. Trump stood at 31.1 percent in our Iowa polling average on the night of the debate, so if he finishes at 27.1 percent, he’ll have lost 4 percentage points since then.

I spoke about this in a recent post: Candidates with late momentum in the polls and candidates who appeal to Christian conservatives have tended to outperform their polling. The latters seems to be holding true of the entrance polls as well. Cruz is doing better than his early entrance poll numbers suggested, and Rick Santorum did the same in 2012.

Harry, so it seems like that first wave of exit polls was — as expected — a bit off: Cruz looks more competitive?

I should point out that Sanders is closing as more votes have been reported. Clinton is up 51.8 percent to Sanders 47.6 percent. That’s with 34 percent of precincts reporting. And literally as I refresh it’s 36 percent of precincts and now it’s 51.5 percent to 48 percent.

https://twitter.com/FiveThirtyEight/status/694340786272350209

Jeb Bush is currently projected to win about 2,600 votes in Iowa, or about 2 percent of the overall turnout. That would work out to one vote per $25,000 in spending by Bush’s super PAC, Right to Rise, which has spent a total of $64.8 million so far. (Granted, not all of that money has been spent in Iowa.)

Harry, I’m going to go out on a limb and guess that Sen. Tim Scott didn’t just wake up today and decide to throw his support Rubio’s way. For Team Marco, tonight seems like it’s all about vaulting from laying low/tamping down expectations to pouring it on and claiming “big mo.” One wonders who else is waiting in the wings.

So far with 7 percent of precincts in, Cruz leading 30 percent to 29 percent for Trump, 18 percent for Rubio. Raw votes so far much closer between Trump/Cruz than entrance polls indicated, and Rubio further down. BUT, this could be misleading: These are most likely tiny, more rural caucus sites that are easier to count fast. For Trump and Rubio, bigger is better. As YUGE suburban precincts start reporting votes, I’d expect Trump’s and Rubio’s numbers to go up.

I wrote last week about LULAC Iowa’s campaign to get 10,000 Latinos to caucus tonight. If they’re successful, Latinos will be about 5 percent of caucus-goers, while only making up 2.9 percent of the electorate. One of the big questions, though, was which candidate would benefit from a higher turnout. We won’t know what share of caucus-goers are Latino for another month, but we do know which counties have the highest share of Latinos. In 2014, 27.3 percent of Crawford County residents were Latino — the most of any county in Iowa. There are eight precincts in Crawford County, and we have results for only one of them so far. But it might give us a glimpse at who Latinos are voting for: The breakdown is 50 percent for Clinton, 33.3 percent for O’Malley and 16.7 percent for Sanders.

I was just looking at Clinton’s numbers in 2008, and I was struck by how even the three candidates (Obama, Clinton, Edwards) were in so many counties. There were lots in which all three got between 25 percent and 35 percent of the vote. What that really indicates is that Edwards was a significantly stronger candidate there than, say, his current status would suggest.

Rubio is going to pick up the endorsement of Sen. Tim Scott of South Carolina. That’s a big one for him.

It’s early on the Democratic side, but interesting that Hillary Clinton is leading in considerably more counties than Bernie Sanders: 34 of them to Sanders’s 23, with 10 percent of precincts reporting. That could matter because Democratic delegate allocation rules reward candidates who perform well throughout Iowa instead of having their vote concentrated in a few large precincts.

As Dave said, the split across region in vote is pretty much nil on the Democratic side. It could be why Clinton is holding a consistent ~7 percentage point lead over Sanders in state equivalent delegates with now 14 percent of precincts reporting.

One last thing about the mailers: You may have seen this (there was a Mother Jones story) — one of the people who got a mailer that incorrectly reported his voting habits was Iowa State political scientist Dave Peterson, who is editor of the journal Political Behavior, and (at least as importantly) my friend and collaborator. He wasn’t too thrilled about being “defamed.”

Regionally, the split on the Democratic side looks remarkably even, with Clinton leading Sanders modestly in all areas of the state. On the GOP side, there’s much more variation: As expected, Rubio is doing best in “Eastern Cities” and Des Moines. But the surprise is that Cruz’s support (at least based on the entrance polls) appears to be fairly even across regions, while Trump’s skews toward rural and western Iowa. Maybe Trump’s stronger-than-expected performance with evangelicals has something to do with that? We’ll see when more actual votes come in.

https://twitter.com/jodyavirgan/status/694335092869709824

In a testament to the where-there’s-smoke-there’s-fire power of Twitter, only a few minutes into the caucus process, the rumors were flying that Ben Carson would be dropping out of the race. This was precipitated by CNN reports that Carson would not be continuing straight onto New Hampshire like other candidates but would instead be going to his home in Florida and then to Washington, D.C., for the National Prayer Breakfast (scene of his 2013 breakout political moment). Via a couple of tweets, Carson spokesman Jason Osbourne said the candidate was not suspending his campaign, but rather getting fresh clothes. Osbourne further noted in a direct response to CNN’s Jake Tapper that the good doctor had been on the road for 17 days, leading one to surmise that he does not believe in doing his laundry in the hotel sink. You can’t blame people for thinking that Carson is going to be on his way out soon, though; his polling has been anemic, in the single digits for the past few weeks, and it’s hard to fathom that he’ll stay in this race too much longer.

I should note we have actual statewide delegate equivalents on the Democratic side: So far, with 6 percent in, it’s Clinton 52.5 percent to Sanders 46.6 percent.

The weeks leading up to this year’s Iowa caucuses brought America debates, media coverage and the march of indefatigable candidates across the state. But it also brought a torrent of donations to super PACs for candidates on both sides of the aisle. Among them: George Soros donated $6 million to Priorities USA Action, a super PAC trying to help Clinton. And investor Ken Griffin gave $2.5 million to Conservative Solutions, which is working to elect Rubio. Both donations were made in December and disclosed as part of financial reports to the Federal Election Commission.

So I mentioned earlier that I think there are different things going on here for the two parties. For the Republicans, I think one of the implications is that if Rubio looks anything remotely like a viable nominee at the end of the night, this will have reordered the race for a lot of people who’ve been expecting him to do well throughout the invisible primary. We haven’t had too many hard indicators that his candidacy has taken off. But doing well — loosely defined — in Iowa would be one.

I’ve seen some good discussion online about whether it’s fair to characterize Bernie Sanders’s support as being confined to more liberal Democrats. This is important for Sanders because Democratic electorates in states beyond Iowa and New Hampshire are not as liberal as the ones there. So far, however, it appears that Sanders’s support is coming from exactly who you might expect. According to the Iowa entrance poll, Sanders leads Clinton by 19 points among “very liberal” voters but trails her by 6 points among “somewhat liberal” voters and by 28 points among moderates.

Let me add to what David said: Rubio is also leading among “somewhat conservatives” in the early entrance polls with 28 percent to Trump’s 27 percent. The candidate who does well among somewhat conservatives tends to win the nomination.

One of the most promising signs in the entrance poll for Rubio is that he’s leading with college graduates: 28 percent Rubio, 25 percent Trump, 20 percent Cruz. Also with post-graduate degree-holders: 25 percent Rubio, 24 percent Cruz, 18 percent Trump. If Rubio finishes above 20 percent as entrance polls suggest, this is why.

My best guess is that someone who regularly votes/participates but was thinking about sitting out this year for whatever reason might have been susceptible to the shame appeal, under the best of circumstances. But as Dave points out, the overall presentation really seems to have missed the mark. As did the fact that the voter “grades” were not accurate.

There’s certainly strong evidence that social pressure and past turnout records have been effective motivators, but my impression is that Cruz badly blew it here: There was no need to use the word “violation” and make it seem like voters were getting a photo enforcement speeding ticket. Of course, we’ll never truly be able to gauge its impact. It’s possible a few voters who had a hunch they didn’t like Cruz’s personality found last-minute reinforcement from either the mailer or the stories about it.

Any thoughts on whether that “voting violation” mailer from Cruz will have any effect, positive or negative? It seems to be based on some famous studies on the effect of social pressure on turnout, but it was pretty over-the-top.

We’ll see if the entrance polls on the Democratic side are anywhere close to the actual result, but Clinton holds a clear lead in them right now, 51 percent to 42 percent. Keep in mind that the statewide raw vote could underestimate her strength in state delegate equivalents.

Entrance polls have Donald Trump narrowly leading on the Republican side, although remember that some voters make up their minds (or can have their minds changed) at the caucus site. They also suggest, however, that Marco Rubio made up a lot of ground on Trump at the last minute. Among voters who decided “in the last few days,” Rubio won 29 percent of the vote to just 10 percent for Trump. And Rubio, Trump and Ted Cruz were virtually tied among voters who said they made their minds up just today.

Harry, that is a wild stat! But it kinda makes sense: The 17-to-29-year-olds who are likely to turn up at something like this are likely to be the most committed young partisans. It’s not an obvious leap that they’d be on fire for Sanders, but it seems pretty plausible. The sort of young folks who read up on the issues and are mobilized by young activists.

A: I’d say we have a pretty good idea of which counties aren’t swing counties tonight. On the Democratic side, we expect Sanders to win Johnson County (Iowa City) and Poweshiek County (Grinnell), while we’d expect Clinton to perform well in areas of the state with older voters like Wapello County (Ottumwa) or Des Moines County (Burlington). On the Republican side, we’d expect Cruz to do well in the heavily Dutch/evangelical counties (Sioux, Mahaska, Plymouth, Warren) and Rubio to do well in more secular suburban counties (Dallas, Polk, Scott). OTOH, which are the bellwethers? On the Democratic side, I’ll be looking at Cerro Gordo (Mason City) and Dubuque counties, which are very liberal but full of working-class whites. On the Republican side, a good bellwether might be a mid-size town like Fort Dodge (Webster County) or Marshalltown (Marshall County). CORRECTION (Feb. 1, 9:10 p.m.): A previous version of this post incorrectly listed a mid-size town in Webster County; it is Fort Dodge, not Dodge City.

A lot of readers have asked where they can get tonight’s results as they come in. This year, both the Iowa Republican and Democratic parties have partnered with Microsoft to develop a new reporting app — you can find that data directly here for the Republican side and here for the Democratic side. In the past, results were reported from the precincts by dialing in to each party’s headquarters and using a phone keypad to input the votes — not exactly the best system for error-free data. As I’ve written, even the app won’t guarantee a perfectly precise count, but hopefully, it’ll mean that we can get verified results a little more quickly than we have in the past.

Here’s an absolutely nutso stat from the early entrance poll: 17-to-29-year-olds on the Democratic side are going 91 percent to 8 percent for Sanders over Clinton. They make up only 15 percent of the electorate, however.

Entrance polls are reporting a slight lead for Hillary Clinton among Democratic voters in Iowa. But there are two things to keep in mind, above and beyond the fact that entrance and exit polls can sometimes be inaccurate. First, Iowa Democrats can change their vote at the precinct. In particular, supporters of candidates who have less than 15 percent in their precinct can be recruited to throw their support to other candidates. In most precincts, that means Martin O’Malley supporters will have to choose between Clinton and Bernie Sanders. Some polls suggest that more of them will back Sanders, which could be worth a percentage point or two for him. But second, Iowa Democrats don’t actually count votes. Instead, they count “state delegate equivalents.” The way the math works on this is that it’s better to have your voters dispersed throughout the state than concentrated in a few large precincts — potentially a problem for Sanders because his vote is thought to be oversubscribed in college towns and urban areas.

It’s entrance poll data, so I wouldn’t draw firm conclusions, but the one thing I might say is that the Rubio surge may, in fact, be real. But, again, early entrance polls are not anything I’d bet my life on. Not at all.

Harry, entrance poll data is rocketing around Twitter — do we know anything yet?

As of this writing, 17 people in the comments section have asked, “What time can we expect results tonight?” Good question! According to our politics editor Micah Cohen, who apparently has time to go look this stuff up, here’s when the results were reported in 2012 (all times Eastern) 9 p.m. — 5 percent reported 10 p.m. — 25 percent reported 10:30 p.m. — 50 percent reported 11 p.m. — 90 percent reported Midnight — 96 percent reported So, that’s some context. We’ll likely know between 10:30 and 11 p.m. Keep in mind that this is the first year that results are being reported by an app, which may speed things up a bit. But it’s gonna be a long night. Chug a coffee.

https://twitter.com/ForecasterEnten/status/694325858136395778

It’s become political cliché to say Iowans cherish their state’s status as the first to cast votes in the presidential campaign. It turns out that Ann Selzer, the Iowa pollster with an illustrious track record whom my colleague Clare profiled last week, has asked Iowans how they feel about their status several times over the decades. And most of them really do cherish being first — not just for Iowa’s sake, but for the nation’s. In the Des Moines Register’s Iowa poll a year ago, 69 percent of adult Iowans said that it’s best for the country if Iowa keeps going first, according to Michelle Yeoman, research assistant and office manager at Selzer & Co. Just 13 percent said it’d be best if some other state went first. And back in February 2008, just after that year’s caucuses, a similar question got almost the exact same result: 66 percent said it’s best that Iowa goes first, and 13 percent disagreed. In January 2000, just before that year’s caucuses, only a quarter of likely caucus-goers said that Iowans don’t reflect the nation as a whole (an NPR analysis suggests they do better than most states). And in 1987, before Selzer was running the Iowa poll, just 7 percent of Republican and Democratic caucus-goers wanted Iowa to let another state go first. Most Iowans do have one big complaint with the caucuses, though: Not enough people participate. In 2000, 57 percent of likely Democratic caucus-goers and 61 percent of likely Republican caucus-goers agreed that “too few Iowans actually get involved in the process and attend their precinct caucus” and said it’s a major problem.

Back to the ground game: We have a pretty good idea that in-person mobilization works to get people to the polls. You can do a field experiment where you compare block by block. But it’s much more complicated to untangle the impact of mobilization strategies on a particular election, with other complex factors in play.

With the exception of a few weeks in which terrorism news dominated, most of the campaign so far — on both the Republican and Democratic side — has been dominated by economic issues, both explicit and implicit. Bernie Sanders, of course, rails against “millionaires and billionaires” at every campaign stop. Donald Trump wants to kick out undocumented immigrants and get tough on China. And both parties’ debates have featured obligatory paeans to the disappearing middle class. But voters going to caucus sites in Iowa this evening are experiencing an economy that is, by and large, pretty good. The state’s unemployment rate is 3.4 percent, one of the lowest in the country and lower than it was when the national recession hit eight years ago. Wages are rising significantly faster than inflation, and overall economic output has fully recovered from what was, in Iowa, a comparatively mild recession. Yet when I visited Iowa back in the fall, I heard much of the same economic anxiety that voters are expressing around the country. Iowa is an extreme example of a central paradox in this election: The economy is better, but voters don’t seem to believe it. Except, maybe it isn’t a paradox at all. As I wrote Friday, there’s a growing disconnect between how Americans see their short-term and their long-term economic prospects. Right now, unemployment is falling and the economy is doing pretty well (a tumultuous start to the year in the stock market notwithstanding). But over the longer term, household incomes have been more or less stagnant for more than 15 years. Young Americans are grappling with student debt; older ones are worried about saving for retirement; parents are worried about their children’s chances of finding good jobs. They’re looking for a president who can address those longer-term problems. Last fall, I met Mary Britton, an office manager in Davenport, Iowa. She had been a “grandmother for Obama” back in 2008 but thought John Kasich might provide the best fix for an economy she feared no longer offered opportunities for the next generation. “All the jobs that allowed middle-class people to prosper, a lot of those jobs went away,” Britton said at the time. “A whole bunch of people are being left behind.”

You’ll no doubt hear a lot about the “ground game” tonight. And much of that talk will mention the role of data in targeting voters. It will be interesting to see how much the hype bears out, but when I visited field offices in Iowa last month to get a sense of how the campaigns were putting data to work, one thing that struck me was how much the particular rules of the Iowa caucuses affect the way data is employed. On the Democratic side, where one’s vote is declared in public and persuasion is baked into the system, that means the voter outreach is given a more personal touch. At Clinton field offices, the same volunteer often called the same voter multiple times. Precinct captains are expected to know the second choice of their area’s voters, should they have to convince them to switch sides. In the GOP race, voter targeting feels a little more geared towards efficiency. The goal is to just drive as many people to the polls as possible. Iowa’s quirky rules can be tough to get your head around, but they really do set the tone for the entire campaign.