Nate Silver's Blog, page 135

Clinton Is Winning The States That Look Like The Democratic Party

“Secretary Clinton cleaned our clock in the Deep South, no question about it,” Bernie Sanders said during Thursday night’s Democratic debate in Brooklyn. “That is the most conservative part of this great country,” he continued. “But you know what, we’re out of the Deep South now. And we’re moving up.”

I have a few problems with this line of argument, which seems to imply that Democratic voters in the Deep South don’t reflect the larger Democratic electorate. (The remarks Thursday night echo previous comments made by Sanders and his campaign.) Consider Sanders’s reference to the term “Deep South,” which traditionally describes Alabama, Georgia, Louisiana, Mississippi and South Carolina: These are five of the only six states, along with Maryland, where at least a quarter of the population is black. Given the United States’ history of disenfranchising black voters — not to mention the importance of black voters to Democrats in November — it’s dicey for Sanders to diminish Clinton’s wins there.

But the Deep South isn’t Sanders’s only issue. His problems in the rest of the South are what really dooms him. Clinton’s largest net delegate gains over Sanders came from Texas (+72) and Florida (+68), two states that are within the South as the Census Bureau (and most other people) define it. Clinton also cleaned Sanders’s clock in Virginia and North Carolina. Overall, Clinton gained a net of 155 delegates on Sanders in the five Deep South states, but she also added 211 delegates to her margin in the rest of the region.

DELEGATE COUNTREGIONCLINTONSANDERSNETDeep South*22570Clinton +155Other Southern states493282Clinton +211Rest of country589745Sanders +156Overall1,3071,097Clinton +210Clinton has dominated Sanders in the South

* Alabama, Georgia, Louisiana, Mississippi and South Carolina

Source: FiveThirtyEight Delegate Tracker, The Green Papers

In addition to being important to the Democratic Party’s electoral present and future, Florida, Virginia, North Carolina and Texas are quite diverse. They’re diverse ideologically — Miami and Austin aren’t exactly “the most conservative part” of the country — and they’re diverse racially. They contain not only a substantial number of African-Americans but also Hispanics and, increasingly, Asian-American voters.

In fact, these states are among the most demographically representative of the diverse Obama coalition that Clinton or Sanders will have to rely on in November.

Although it will be a couple of decades before the electorate as a whole is majority-minority, the Democratic vote is already getting there. In 2012, only 55 percent of President Obama’s voters were white, according to the national exit poll. Our demographic projections of this November’s electorate, which account for population growth since 2012, calculate that the white share of the Democratic vote will tick down another percentage point, to 54 percent. The rest of the Democratic vote will be black (24 percent), Hispanic (15 percent), or belong to Asian or other races (7 percent), according to our projections.

So let’s take those projections as being maximally representative of the broader Democratic electorate as it stands today. In which primary or caucus states has turnout come closest to those ratios?

In 21 states to have voted so far, we have data on this from exit polls. See here for Virginia, for example, where Democratic turnout was 63 percent white, 26 percent black, 7 percent Hispanic and 5 percent Asian or other when it voted on Super Tuesday. That’s pretty close to the Democratic electorate overall, although with too few Hispanic voters. In the other 29 states — those that haven’t voted yet or where no exit poll was conducted — I’ll estimate the Democratic electorate based on our demographic projections, with an adjustment for the fact that the Democrats who vote in primaries are somewhat whiter than those who vote in November.1

Then I’ll calculate the root-mean-square error (RMSE) for each state — a measure of the difference between the demographics of its primary or caucus turnout and the projected Democratic electorate in November. A lower RMSE is better for our purposes, because it means the state’s demographics are more representative of the national Democratic coalition.

DEMOCRATIC PRIMARY/CAUCUS ELECTORATESTATEWHITEBLACKHISP./ LATINOASIAN/ OTHERRMSE*OUTCOMENew Jersey57%26%11%6%5%Illinois5828958Clinton +2Florida48282059Clinton +31New York61201639Virginia63267512Clinton +29Nevada591319813Clinton +5North Carolina62313416Clinton +14Maryland47398618Tennessee63322318Clinton +34Arkansas67273318Clinton +37Michigan68233618Sanders +2Pennsylvania70198319California4911231719Delaware69243320Texas431932721Clinton +32Arizona62725621Clinton +15Kentucky72214422Missouri72213422Clinton +0Connecticut74157324Ohio74212324Clinton +14Alaska665111725Sanders +59Oklahoma74144925Sanders +10Indiana76138326Colorado72416827Sanders +17Kansas78107529Sanders +36Washington744101229Sanders +46District of Columbia45494229Utah76411930Sanders +57Rhode Island7998430Georgia38517333Clinton +43Wyoming80410633Sanders +11New Mexico453391334Idaho80311534Sanders +57Minnesota8293634Sanders +23Wisconsin83103434Sanders +14Montana8246936Alabama40541536Clinton +59South Dakota82361036Nebraska8583337Sanders +14Massachusetts8546538Clinton +1North Dakota8644639West Virginia8855340Oregon8714742South Carolina35612244Clinton +47Iowa9134244Clinton +0Hawaii45284646Sanders +43New Hampshire9321447Sanders +22Louisiana26638249Clinton +48Maine9521249Sanders +29Vermont9510350Sanders +72Mississippi24711458Clinton +66Projected Democratic electorate in November5424157—

* Root-mean-square error of turnout as compared with projected Democratic electorate in November

Source: Exit polls, FiveThirtyEight demographic projections

The most representative state by this measure is New Jersey. We expect its primary electorate to be about 57 percent white, 26 percent black, 11 percent Hispanic and 6 percent Asian or other, quite close to the national Democratic electorate. New Jersey won’t vote until June 7, although Clinton was well ahead when the last poll was released there in February.

After New Jersey comes Illinois, which Clinton won narrowly — and then Florida, where Clinton won going away. Then there’s New York, which votes Tuesday, and where Clinton is 15 percentage points ahead in our polling average. Virginia, another Southern state, ranks as the next most representative; Clinton won it easily. Then there’s Nevada, another Clinton state, before we go back to the South to North Carolina, also won by Clinton. The next group of four states (Maryland, Tennessee, Arkansas and Michigan) are roughly tied and include some further representation for the South, along with, finally, one state (Michigan) that Sanders won.

In other words, Clinton has won or is favored to win almost every state where the turnout demographics strongly resemble those of Democrats as a whole. This shouldn’t be surprising — Clinton is winning nationally by about 14 percentage points in the popular vote. So if you’re in a state that’s well-representative of Democrats’ national demographics, you might expect her to win it by a solid margin too.

Listen to the latest episode of the FiveThirtyEight elections podcast.

http://c.espnradio.com/s:5L8r1/audio/2729218/fivethirtyeightelections_2016-04-15-132918.64k.mp3?ad_params=zones%3DPreroll%2CPreroll2%2CMidroll%2CMidroll2%2CMidroll3%2CMidroll4%2CMidroll5%2CMidroll6%2CPostroll%2CPostroll2%7Cstation_id%3D4278Subscribe: iTunes |Download |RSS |Video

It’s true that a couple of states in the Deep South, such as Mississippi and Louisiana, rate as being not well-representative by our definition. But overall, that’s more likely to be true of places where Sanders has won. New Hampshire ranks as the 46th most representative out of 50 states (and Iowa ranks 44th — maybe those states shouldn’t hold the first two contests?). Wisconsin, which Sanders won last week, is below average.

And the sort of wishful thinking Sanders is engaged in can cut both ways. Yes, Clinton’s lead would be considerably narrower (although she’d still be winning) without delegates from the Deep South. But what if you excluded delegates from caucuses, where Sanders has gained a net of 150 delegates on Clinton? Without those delegates, Sanders couldn’t even maintain the pretense of a competitive race. Not only are most of those caucus states extremely white and therefore poorly representative of Democrats’ national demographics — many of them (such as Idaho and Nebraska) are also quite red. Furthermore, caucuses tend to disenfranchise voters by making it harder to vote. Our demographic modeling suggests that this has hurt Clinton and that Sanders wouldn’t have won by the same enormous margins if those caucus states had held primaries instead.

But overall, the math is pretty simple. Sanders is winning states that are much whiter than the Democratic electorate as a whole, Clinton is winning states that are much blacker than the Democratic electorate as a whole, and Clinton is winning most of those states that are somewhere in the middle, whether they’re in the South (like Virginia) or elsewhere (like Ohio or Nevada). That’s why she’ll probably be the Democratic nominee.

A State-By-State Roadmap For The Rest Of The Republican Primary

Three weeks ago, when we last took a detailed look at Donald Trump’s quest to win 1,237 delegates, his path looked rocky but endurable. The panel of eight experts FiveThirtyEight assembled projected Trump to wind up with 1,208 by the time California and four other states finished counting their votes on June 7, a number that would leave him tantalizingly close to clinching the Republican presidential nomination — probably close enough that he’d be able to get over the hump by persuading some uncommitted delegates to come his way before the convention.

Since then, Trump has gotten mostly bad news. Last week, he lost Wisconsin, which our panel originally considered a toss-up state leaning in Trump’s direction. Then this weekend, he was shut out of delegates at the Colorado state convention. He’s also had a couple of minor setbacks; Trump got no delegates in Utah when we thought he might get a few. All told, Trump would finish with 1,175 delegates if he hits our panel’s original estimate in the remaining states. That’s far enough away from 1,237 that winning over enough uncommitted voters will be challenging, especially given Trump’s lack of success at finding pro-Trump delegates.

But we also have a lot of new information at our disposal. Trump’s polling has held up well in the Northeast, and he has a good chance to beat the panel’s original projections in New York and Connecticut. On the flip side, his loss in Wisconsin bodes poorly for his performance in Indiana, another state we originally had as leaning toward Trump. So it’s time to revisit our projections, going through the remaining states one at a time. (Given how much delegate rules vary from state to state, there’s really no avoiding this level of detail, much to the bane of my editor.)

This time, it’ll just be me (Nate) instead of the panel, but I’ll aim to compensate by issuing three types of projections for each remaining state:

First, a deterministic projection, which lists the single most likely outcome in every state, in my view. Say there’s a state with 50 winner-take-all delegates, and we give Trump a 60 percent chance of winning it. Since the win is more likely than not, we’d score all 50 delegates in his column.Second, a probabilistic projection, which hedges its bets. In the aforementioned example, Trump would have a 60 percent chance of winning 50 delegates and a 40 percent chance of winning no delegates, which works out to an expectation of 30 delegates.Finally, a path-to-1,237 projection, which is not the same thing as Trump’s best-case scenario. Instead, it’s roughly what I consider his path of least resistance to finish with 1,237 delegates exactly, not counting uncommitted delegates.1 It’s up to you, dear reader, to think about how realistic it might be, or not — I’ll chime in with some thoughts at the end.

Where they’re available, I’ll use FiveThirtyEight’s poll-based forecasts to inform these projections. (Specifically, I’ll use an average of the “polls-only” and “polls-plus” versions of our model.) However, there isn’t a lot of recent polling in some states, so instead I’ll make inferences from Trump’s performance in nearby states, the state’s demographics, and so forth, tending to give a fair amount of deference to the panel’s original forecasts.

The first stop on our tour is Cheyenne, Wyoming, where Republicans will hold their state convention this weekend.

Wyoming

Saturday, April 16
Bound delegates available: 14
Original Trump delegate projection: 1
Allocation method: State convention

Wyoming is one of three states, along with Colorado and North Dakota, where Republicans don’t hold a presidential preference vote. Instead, Republican voters elect delegates through a series of local and state conventions. So far, such conventions have gone terribly for Trump in other states. He got shut out of delegates in Colorado, and while delegates aren’t officially bound to anyone in North Dakota,2 the overwhelming majority chosen there are likely to prefer Ted Cruz.

Wyoming isn’t really going any better for Trump. It’s already halfway through its process, having elected 12 delegates — including just one for Trump — at county conventions on March 12. (We don’t count these March delegates in the totals listed here.) If anything, Trump was lucky to get even one delegate given how lopsided the vote was in most counties, meaning that a shutout is possible when the state convention picks 14 more delegates on Saturday. Revised Trump delegate projections: deterministic 0; probabilistic 1; path-to-1,237 1.

New York

Tuesday, April 19
Bound delegates available: 95
Original Trump delegate projection: 71
Allocation method:

14 delegates awarded proportionally based on statewide vote, with 50 percent winner-take-all trigger.Three delegates in each congressional district (81 total), split 2-1 between top two finishers, with 50 percent winner-take-all trigger.

Ahh, New York. So often the center of attention in everything else, now the center of attention in presidential politics, too. Politicians — they’re just like us! We get to see them awkwardly pandering to local customs and stuffing their face with fatty food.

But I digress. The headline is that Trump is very likely to beat our original projection of 71 delegates in New York. The mechanics of this are a little complicated to work out, however. I’m going to go through New York in some detail because many of the same principles will apply in California and other states.

The delegate rules themselves are tricky in New York. Fourteen delegates are awarded based on statewide results; they’re allocated proportionally,3 but they become winner-take-all if a candidate achieves at least 50 percent of the vote, a threshold that Trump exceeds (although barely in most cases) in almost every recent poll. The other 81 delegates are awarded, three at a time, in each of New York’s 27 congressional districts. In most cases a candidate gets two of three delegates for winning a district, but he gets all three if he wins the district with a majority of the vote.4

I’m not going to overthink Trump’s 50-percent-plus standing in statewide polling: He has yet to get a majority anywhere else, but New York is his home state. The more difficult question is how this translates to congressional districts; if Trump winds up with (for instance) 54 percent of the vote statewide, in how many districts will that translate to a majority?

You should be wary of quick-and-dirty answers from polls that break out the results by region. For one thing, the sample sizes on those regional breakouts are usually quite small. But also, the precise way the regions are defined is important; they may not tell us very much about the deep-blue districts we’re interested in.

One recent poll, for instance, broke out results from “New York City and Long Island” and showed Trump beating his statewide numbers there. Here’s the problem with that: New York City is overwhelmingly Democratic, while Long Island5 has a fair number of Republicans. (To put things in perspective: There are 11 times more Republicans in the Long Island-based 4th Congressional District than there are in the 15th Congressional District, in the South Bronx.) Among the registered Republicans who live in either Long Island or New York City, about 60 percent are in Long Island. Thus, Long Island will make up a majority of the sample in a poll of Republicans in New York and Long Island. However, Long Island has only four congressional districts — as compared with New York City’s 11 or 12 — and therefore a relatively small minority of the delegates at stake.

The ways around this are to take an extremely large sample so that you can actually pinpoint those blue districts, or to do some sort of statistical modeling. One pollster, Optimus, took the former approach, surveying more than 14,000 (!) New Yorkers in a “robocall” survey and providing a breakdown of results by congressional district. That poll had Trump finishing first in 26 of 27 districts (losing one Manhattan district to John Kasich) and getting a majority in 15 of 27, along with a majority of the vote (just barely) statewide. Such a result would earn Trump 82 of 95 delegates.

Note, however — especially since this has implications for other states — that the Optimus poll did not have Trump doing especially well in extremely Democratic districts in Manhattan, the Bronx, Brooklyn and Queens. Instead, he performed extremely well in Long Island, where the poll had him with 57 percent of the vote — along with the Republican-leaning NYC borough of Staten Island, where he had 68 percent.

The poll also showed a lot of variance in Trump’s vote from district to district in New York. Therefore, even if Trump gets 55 percent or 60 percent of the vote statewide, he’ll probably be under 50 percent in a handful of districts, preventing a clean sweep.

But unless Trump has trouble getting his voters to the polls because of New York’s strict registration laws, he doesn’t have a lot to worry about. He should beat our original estimate of 71 delegates unless he finishes below 50 percent statewide. Given the recent polls, our models calculate there’s less than a 10 percent chance of his failing to hit that majority threshold. Revised Trump delegate projections: deterministic 85; probabilistic 83; path-to-1,237 91.

Connecticut

Tuesday, April 26
Bound delegates available: 28
Original Trump delegate projection: 19
Allocation method:

13 delegates awarded proportionally based on statewide vote, with 50 percent winner-take-all trigger.Three delegates in each congressional district (15 total) awarded winner-take-all based on district vote.

Five Northeastern states vote on April 26. The most important is probably Pennsylvania, for reasons I’ll explain in a moment. But Pennsylvania’s delegate rules have a complicated twist, so let’s start with Connecticut, another state where Trump has a chance to beat our original projections.

Connecticut’s statewide delegates are awarded proportionally, but (as in New York) they become winner-take-all for a candidate getting a majority of the statewide vote. So, does Trump have a shot of getting to 50 percent? According to the only recent poll of the state, from Emerson College, he certainly does; that survey has him with 50 percent of the vote exactly. Our polling average in Connecticut rounds Trump down to 48 percent because Emerson’s polls have had a Trump-leaning house effect, but it’s obviously a close call as we await more polling.6

And there’s some further good news for Trump. Connecticut awards three winner-take-all delegates by congressional district. Because its congressional districts are fairly homogenous — especially relative to New York’s diverse ones — they’ll probably be swept by anyone who dominates the statewide vote, as Trump looks like he will.

In some sense, those results would speak to the primacy of geography over demography in the GOP primary. Demographically, Connecticut is a lot more white collar than the areas where Trump has excelled. But Trump got almost 50 percent of the vote in neighboring Massachusetts, and it looks like he’ll get there in New York also. Revised Trump delegate projections: deterministic 23; probabilistic 24; path-to-1,237 28.

Listen to the latest episode of the FiveThirtyEight elections podcast.

http://c.espnradio.com/s:5L8r1/audio/2725535/fivethirtyeightelections_2016-04-11-183435.64k.mp3?ad_params=zones%3DPreroll%2CPreroll2%2CMidroll%2CMidroll2%2CMidroll3%2CMidroll4%2CMidroll5%2CMidroll6%2CPostroll%2CPostroll2%7Cstation_id%3D4278Subscribe: iTunes |Download |RSS |VideoMaryland

Tuesday, April 26
Bound delegates available: 38
Original Trump delegate projection: 31
Allocation method:

14 delegates awarded winner-take-all based on statewide vote.Three delegates in each congressional district (24 total) awarded winner-take-all based on district vote.

Trump is an 86 percent favorite to win Maryland according to our polls-only forecast and has a 71 percent chance according to polls-plus. That range of odds sounds about right to me. Trump did not do especially well in neighboring Virginia, where he won with only 35 percent of the vote. But whereas Marco Rubio emerged as the principal alternative to Trump in Virginia, it’s not clear whether Kasich or Cruz is the more viable non-Trump in Maryland; they’re roughly tied for second place in our polling average. That could make it hard for the anti-Trump vote to consolidate.

Kasich and Cruz could have more success at denying Trump delegates at the congressional district level. (Maryland awards three delegates to the winner of each district.) Trump did poorly in the Washington, D.C., suburbs on Super Tuesday — winning, respectively, just 22 percent, 26 percent and 29 percent of the vote in Virginia’s 8th, 11th and 10th congressional districts. Kasich is perhaps an outright favorite over Trump in Maryland’s suburban 4th Congressional District, with the 5th and 3rd potentially competitive. Cruz is probably Trump’s main competition in the 6th Congressional District in Maryland’s Appalachian panhandle and the 1st Congressional District on its Eastern Shore, although Trump will probably win both — perhaps by huge margins, given his performance in similar areas in Virginia.

With such diverse districts, a clean delegate sweep for Trump would be hard to pull off. Winning six of eight districts (along with the statewide vote) would constitute a good day for him and would win him 32 delegates, just slightly ahead of our original estimate of 31. In the absence of detailed polling, I’d probably bet on Trump’s getting five of the eight districts instead, which would be worth 29 delegates. However, there’s downside for Trump if he loses the statewide vote, which can’t totally be ruled out, according to the polls. Revised Trump delegate projections: deterministic 29; probabilistic 27; path-to-1,237 32.

Delaware

Tuesday, April 26
Bound delegates available: 16
Original Trump delegate projection: 15
Allocation method: Winner-take-all based on statewide vote.

One more simple state before we get to Pennsylvania. Delaware awards 16 delegates to the statewide primary winner. The words “we could really use a poll of Delaware” have apparently never been written before on the internet, according to Google. But, we could really use a poll of Delaware! Until then, I’ll assume Trump is favored based on his polling leads in Pennsylvania and Maryland. Delaware’s votes are surprisingly high-leverage — based on the ratio of voters to delegates, a Republican vote in Delaware is worth about four times more than one in Florida, which voted on March 15. So it will be interesting to see if one of the candidates — perhaps Kasich, who has some endorsements from local Republican leaders — will make a point of campaigning in the state. Revised Trump delegate projections: deterministic 16; probabilistic 13; path-to-1,237 16.

Pennsylvania

Tuesday, April 26
Bound delegates available: 17
Original Trump delegate projection: 16
Allocation method:

17 bound delegates awarded winner-take-all based on statewide vote.Three unbound delegates from each congressional district (54 total) are directly elected on the ballot.

Why does Pennsylvania only have 17 delegates? It doesn’t; it has 71. But only 17 of them are bound by the results of the primary; they’re awarded to the statewide winner. That winner is likely to be Trump, who is an 83 percent favorite according to our polls-plus forecast and 94 percent according to polls-only.

Those probabilities conceal a few differences in the polls, however, with recent surveys showing Trump getting anywhere from 33 percent to 48 percent of the vote. Toward the lower end of that range, Trump is very beatable; toward the upper end, he’s almost certainly not. One thing that helps him, however, is that there’s not a clear second-place candidate, with Cruz and Kasich (who was born in Pennsylvania) tied at 24 percent in our polling average. The results could wind up looking something like Illinois, where Trump won fairly easily with 39 percent of the vote without a clear challenger. Revised Trump delegate projections (bound delegates only): deterministic 17; probabilistic 15; path-to-1,237 17.

The stakes in Pennsylvania are higher than the 17 bound delegates imply, however. Voters will also elect 54 unbound delegates — three from each congressional district — directly on the ballot. See here for a sample of what this looks like: Unlike in other states where delegates are directly elected, such as Illinois and West Virginia, there’s no guidance on the primary ballot as to which candidates those delegates might support.

However, slightly more than half of Pennsylvania’s unbound delegates said they’d support the candidate who wins the popular vote statewide or in their districts, according to a survey conducted by the Pittsburgh Tribune-Review. True, those delegates could change their mind later, with an April primary being a fairly distant memory in the event of a July contested convention. But particularly if Trump (or some other candidate) wins the state emphatically, he could get a couple of dozen unbound Pennsylvania delegates to go along for the ride. At the very least, it’s worth thinking about the bloc of Pennsylvania delegates separately from other categories of unbound delegates; they’ll potentially be more amenable to Trump and are an underrated means by which he might get to 1,237 delegates if he pulls up a bit short after California.

Another question is whether any of the campaigns are capable of directing voters toward their preferred unbound delegates through their ground games, such as by circulating lists in advance of the primary via door fliers or social media. Cruz has pulled off analogous tactics at state conventions, but those involved hundreds of highly informed voters in a captive environment, instead of what should be in excess of a million voters in a primary.

Rhode Island

Tuesday, April 26
Bound delegates available: 19
Original Trump delegate projection: 10
Allocation method:

13 delegates awarded proportionally based on statewide vote.Three delegates in each congressional district (six total) awarded proportionally; they’re split 1-1-1 between the top three finishers unless candidates receive more than two-thirds or less than 10 percent of the vote.

Rhode Island, the final April 26 state, is one of the most proportional on the calendar — unluckily for Trump since its working-class demographics, which resemble some of the areas in Massachusetts where he got a majority of the vote, are strong for him. Three delegates from each of its congressional districts will be split 1-1-1 unless one of the candidates finishes with more than two-thirds of the vote or one of the candidates finishes with less than 10 percent of the vote. Because of this quirk, it could be hard for Trump to meet our original target of 10 delegates even if he has about half the vote statewide, although he should come close. Revised Trump delegate projections: deterministic 9; probabilistic 9; path-to-1,237 10.

Indiana

Tuesday, May 3
Bound delegates available: 57
Original Trump delegate projection: 37
Allocation method:

30 delegates awarded winner-take-all based on statewide vote.Three delegates in each congressional district (27 total) awarded winner-take-all based on district vote.

Indiana, usually an afterthought in electoral politics, played an important role in the 2008 Democratic primary. The state could be a pivot point again this year. It has a date to itself on the primary calendar on May 3 and may end Trump’s winning streak as the Republican contest moves out of the Northeast. Just as important, it has an aggressive delegate allocation method, with 30 delegates awarded to the statewide winner and an additional 27 to the winners of each of nine congressional districts. After California, in fact, Indiana is probably the most important state remaining.

Last month, our panel gave Trump an average of 37 delegates in Indiana, which implied that he’s the favorite to win there. I don’t think I can agree with that after Wisconsin, however. The states are relatively similar demographically. In Indiana, as in Wisconsin, Trump doesn’t have much support from statewide elected officials. Moreover, the Midwest as a whole has been a middling region for Trump. Earlier in the calendar, he got away with some wins in the Midwest with a vote share in the mid-to-high 30s. Now that the field has winnowed and Republican voters have learned to vote tactically, he’ll often need 40 percent of the vote to win a state instead.

Still, some caution is in order. There’s been no polling at all in Indiana. Perhaps Trump can hope that his momentum from April 26 will carry him to victory, or that Kasich will drain a few votes from Cruz in counties that border Ohio. My deterministic projection has Trump losing — although salvaging a few congressional districts — while the probabilistic one is more equivocal. The path-to-1,237 projection has Trump winning, almost out of necessity, because it will be hard for him to carve out a path to 1,237 delegates without the Hoosier State. Revised Trump delegate projections: deterministic 9; probabilistic 22; path-to-1,237 48.

Nebraska

Tuesday, May 10
Bound delegates available: 36
Original Trump delegate projection: 1
Allocation method: Winner-take-all based on statewide vote.

Nebraska has every appearance of being a strong Cruz state. He’s performed well everywhere else in the vicinity, most notably in Kansas, where he beat Trump by 24 percentage points. Yes, most of those earlier results were caucuses — Nebraska holds a primary — but we expect Nebraska to take its place as part of an anti-Trump wall of states that cuts through the prairie. Revised Trump delegate projections: deterministic 0; probabilistic 4; path-to-1,237 0.

West Virginia

Tuesday, May 10
Bound delegates available: 34
Original Trump delegate projection: 33
Allocation method:

Three delegates awarded winner-take-all based on statewide vote.22 statewide delegates, along with three delegates in each congressional district (nine total), are directly elected on the ballot. These delegates may be bound or unbound depending on how they list themselves on the ballot.

Trump has performed well in counties bordering West Virginia. And the economy of the region, heavily affected by outsourced jobs, bodes well for Trump. This is another state where Trump has a good chance to get at least 50 percent of the vote.

The catch is that only three delegates are awarded in West Virginia based on the popular vote. The other 31 are directly elected, in a process that somewhat resembles Pennsylvania’s but with some important differences that increase the potential for chaos.

One difference is that delegates may choose to affiliate themselves with a candidate; if so, the affiliation is listed on the ballot, and the delegates are bound to that candidate at the convention. Other delegates may run and be elected as uncommitted, however.

Another difference is that the ballot is much more disorganized. Voters have to pick not only three delegates from their congressional district but also 22 statewide delegates from a list of 220 candidates!

It’s hard to know how much this might hurt Trump in a state that should otherwise be very good for him, but I suspect that our panel — which originally gave Trump 33 of 34 delegates — may have underestimated the risk. In West Virginia’s 2012 Republican primary, held at a point when the GOP race was no longer competitive, Mitt Romney won 70 percent of the vote. Nevertheless, five of his delegates lost out to uncommitted delegates and another two lost to supporters of Rick Santorum, who had suspended his campaign weeks earlier. There was also a massive undervote, with only about half as many votes cast for at-large delegates as permitted.

A well-organized candidate might be able to take advantage of this in various ways. For instance, he could recruit delegates whose names appeared higher in alphabetical order or Bound delegates available: 28
Original Trump delegate projection: 12
Allocation method: Proportional based on statewide vote.

All right, a simple one for a change. I don’t expect Trump to do well in Oregon — the demographic models we originally drew up in January and February rated it as one of his worst states. But there aren’t a lot of delegates at stake, and they’re allocated highly proportionally. So Oregon just doesn’t matter much. It is interesting mostly insofar as it could give us a few hints about how higher-stakes Washington and California will vote later on. Revised Trump delegate projections: deterministic 11; probabilistic 11; path-to-1,237 12.

Washington

Tuesday, May 24
Bound delegates available: 44
Original Trump delegate projection: 17
Allocation method:

14 delegates awarded proportionally based on statewide vote.Three delegates in each congressional district (30 total) split 2-1 between top two finishers, with 50 percent winner-take-all trigger.8

Washington is not quite as proportional as it’s been billed. Its 14 statewide delegates are allocated basically proportionally.9 But it uses the New York State method for allocating its 30 congressional district delegates, generally splitting them 2-1 in favor of the district winner but giving all three to a candidate who wins 50 percent of the vote or more.

As I mentioned above, demographic models don’t expect Trump to perform well in the Pacific Northwest, but we don’t have many polls, or many results from neighboring states, to confirm that. The safest bet seems to be that given Cruz’s dominance in Idaho panhandle counties that border Washington, Eastern Washington will be problematic for Trump, with Cruz possibly achieving majorities in the 5th and 4th congressional districts. The rest of the state is harder to figure out, but I’m hedging slightly to the downside versus our initial projections. Revised Trump delegate projections: deterministic 15; probabilistic 15; path-to-1,237 18.

New Jersey

Tuesday, June 7
Bound delegates available: 51
Original Trump delegate projection: 51
Allocation method: Winner-take-all based on statewide vote.

Five states vote on June 7, when the big prize is California. I’m going to go at lightning-round speed through the other four; for various reasons, these states aren’t a major source of uncertainty.

New Jersey’s 51 winner-take-all-delegates, for instance, are almost sure to go for Trump, who was well ahead in polls there even before he was endorsed by Gov. Chris Christie.10 Although the New Jersey polling is a bit out of date, Trump’s strong numbers in New York and Connecticut polls bolster our confidence that he’ll win in the Garden State. Revised Trump delegate projections: deterministic 51; probabilistic 48; path-to-1,237 51.

Montana

Tuesday, June 7
Bound delegates available: 27
Original Trump delegate projection: 0
Allocation method: Winner-take-all based on statewide vote.

Montana, by contrast, looks like a safe Cruz state, both according to our demographic model and based on Cruz’s dominant performance in nearby states. Revised Trump delegate projections: deterministic 0; probabilistic 3; path-to-1,237 0.

South Dakota

Tuesday, June 7
Bound delegates available: 29
Original Trump delegate projection: 0
Allocation method: Winner-take-all based on statewide vote.

The same goes for South Dakota, another winner-take-all state where Cruz is a clear favorite. It’s not that it’s impossible to imagine Trump winning South Dakota or Montana — those contests are two months away, and a lot could change. But if things have gotten so good for Trump that he wins South Dakota or so bad for him that he loses New Jersey, the outcome of the Republican contest should be fairly clear by that point. States like these aren’t tipping-point states, in other words. Revised Trump delegate projections: deterministic 0; probabilistic 3; path-to-1,237 0.

New Mexico

Tuesday, June 7
Bound delegates available: 24
Original Trump delegate projection: 10
Allocation method: Proportional based on statewide vote.

New Mexico has only 24 delegates, and they’re allocated proportionally, the only proviso being that a candidate needs at least 15 percent of the statewide vote to qualify. A February poll found Trump running behind his national numbers in New Mexico and trailing Cruz, and our demographic model suggests that New Mexico could be a poor state for Trump. But there isn’t enough at stake for that to matter much. Revised Trump delegate projections: deterministic 9; probabilistic 9; path-to-1,237 10.

California

Tuesday, June 7
Bound delegates available: 172
Original Trump delegate projection: 93
Allocation method:

13 delegates awarded winner-take-all based on statewide vote.Three delegates in each congressional district (159 total) awarded winner-take-all based on district vote.

I suspect the political commentariat hasn’t fully woken up to how monumental a finish California will be. Even if Trump is going gangbusters and meeting or exceeding his path-to-1,237 projections, a poor performance in California could leave him well short of 1,237 delegates. Or if he has a disappointing May, he could salvage a chance at clinching the nomination with a strong finish there.

The math in California is simple enough; 13 delegates go to the statewide winner, but the real prizes are congressional districts, which are worth three delegates each, for 159 total. Given California’s political and demographic diversity, it’s not wrong to think of the state as consisting of 53 micro-primaries. It’s daunting to get a handle on, but this chart might help. It breaks the congressional districts down into some broad categories based on their regions11 and how Democratic or Republican they are:12

As in New York, there are a lot of votes at stake in blue districts. There are nine districts — worth 27 delegates — available in the San Francisco Bay area, all of which are very Democratic. An additional 14 districts, worth 42 delegates, are up for grabs in Los Angeles County, where only 20 percent of eligible voters are registered as Republicans. There are also 10 strongly Republican or Republican-leaning districts in California, however, including several very Republican districts in Orange County.

Who’s going to do well where? I don’t think we have enough data to speak to that reliably. The highly rated Field Poll had Cruz running ahead of Trump in Los Angeles County despite trailing him statewide — big news if true, but as I said about New York, I’ll need larger and more precise samples before coming to very many conclusions. There are also differences of opinion about whether highly blue districts tend to favor Trump overall, disfavor him or something else. Personally, I’m of the view that we don’t really know very much about this and that the rules we learn in one state may not be applicable in others. There’s probably even an element of randomness given that district lines were mostly drawn with general elections in mind and may cut at odd angles across Republican groups of various kinds.

The top-level view in California isn’t all that helpful either, however. Trump leads in our polling average, with 36.5 percent of the vote, but Cruz isn’t far behind, at 30.4 percent. There are a lot of undecided voters, perhaps in part because Rubio voters, who were once fairly plentiful in California, are looking for a new candidate. While our polls-only forecast has Trump with a 65 percent chance of winning California, polls-plus has Cruz favored instead, figuring that Trump is running below his national polls in California (a bearish indicator) and that Cruz has almost two months to figure out how to pull ahead.

Speaking of which, California represents a potential data-mining opportunity for Cruz. Trying to track down Republican voters in East Los Angeles or Nancy Pelosi’s district in California is the equivalent of looking for needles in haystacks. Could his superior data and turnout operations be worth two extra districts? Five? Eight? It’s a factor that could swing the Republican race. For the time being, I’m splitting the difference between our two polling-based models and giving Trump about half the California delegates — close to where we had the state originally. Revised Trump delegate projections: deterministic 94; probabilistic 88; path-to-1,237 112.

So … drumroll please … here’s how our various estimates sum up for Trump:

PROJECTED TRUMP DELEGATESDATECONTESTORIG.DETERM.PROBAB.PATH-TO-1,237April 16Wyoming1011April 19New York71858391April 26Maryland31292732Connecticut19232428Rhode Island109910Pennsylvania16171517Delaware15161316May 3Indiana3792248May 10Nebraska1040West Virginia33292633May 17Oregon12111112May 24Washington17151518June 7California939488112New Jersey51514851South Dakota0030Montana0030New Mexico109910Projected totals417397401479Already won758758758758Total delegates1,1751,1551,1591,237

Although our panel’s original estimates had Trump finishing with 1,175 pledged delegates, my revised deterministic projections have him at 1,155, and the probabilistic version has him at 1,159. I wouldn’t make a huge deal of the differences given the considerable uncertainty in the race, however. Basically, flipping Indiana from a probable win to a probable loss outweighs the gains I have Trump making relative to our original projections in New York and Connecticut. In other states, the differences from the original projections are minor.

At the same time, the path-to-1,237 scenario doesn’t look all that far-fetched — certainly not as compared with, say, Bernie Sanders’s quixotic path to catch Hillary Clinton in pledged delegates. Our path-to-1,237 path has Trump sweeping almost everything in the Northeast, winning Indiana and winning California by a solid-but-not-spectacular margin. I wouldn’t bet on that parlay at even odds, but it’s far from impossible. There’s also a reasonable variation; Trump could win slightly fewer delegates than I’m expecting in New York and Connecticut but make them up with a bigger win in California. So we’re not yet at the point where absolutely everything has to go right for Trump to clinch 1,237 delegates after California; although he can’t afford major setbacks such as losing Indiana or Maryland.

Keep in mind, however, that the question of whether Trump will get 1,237 bound delegates by California is not the same as the question of whether he’ll win the Republican nomination. If he’s close, Trump could still get some uncommitted delegates to come along with him — especially some of the 54 from Pennsylvania if he wins that state. He could also win on the second or later ballot in Cleveland, although I wouldn’t want to count on that if I were in Trump’s shoes.

Do You Have To Be Manly To Be President?

In this week’s politics Slack chat, we’re talking about masculinity and gender in the presidential race. The transcript below has been lightly edited.

simone (Simone Landon, senior editor): So we’ve talked a lot about sexism and how it’s affecting Hillary Clinton in the 2016 presidential race, but this cycle has also seen a lot of, uh, bravado from the male candidates. Who would’ve thought we’d be openly discussing a candidate’s genitalia (I’m of course talking about Donald Trump’s “I guarantee you there’s no problem” in reference to the size of his penis.) What role is masculinity playing on both sides of the race?

clare.malone (Clare Malone, senior political writer): Wooh. A doozy, Simone.

leah (Leah Libresco, news writer: This has been the year of no subtext.

clare.malone: This election, particularly on the Republican side, has been a lot more upfront about sexism, gender and, yes, masculinity (see aforementioned penis). Some of that has to do with Republicans’ spinning the race in order to talk about popular American culture being out of touch with traditional American life — the kind of life/lifestyle they think most Republican voters want. That’s a new spin on the culture wars, but it’s also given the GOP candidates ways to talk about being manly men and to have that characterization make them stand in opposition to what they might see as an increasingly “effeminate” culture — talking about sexism, homophobia, etc., in our everyday lives.

leah: Normally, there are a lot of ways to hint at what a big man you are (i.e., Ted Cruz and carpet bombings, or Jim Webb’s literally having killed a man), and this year, they jumped straight to literal dick measuring.

harry (Harry Enten, senior political writer): Trump has tried to project an image of strength, and while strength needn’t be masculine, I think Trump’s version certainly is. It’s all about physical action. We’re going to bomb, we’re going to build a wall, and so on.

natesilver (Nate Silver, editor in chief): But saying you’re going to bomb countries and build walls is how candidates usually assert their masculinity. Trump’s is different — he’s pure id. Much more explicit.

clare.malone: Right, he’s made the subtext the text.

leah: Some of the enthusiasm for Trump’s brash, crude masculinity seems linked to the way authors like Christina Hoff Sommers argue that masculinity is cramped today, starting in schools — where it’s not just a matter of political correctness, but that school culture, office culture, etc., is all quieter, less physical and less welcoming to a certain subset of men. (Personally, I am also up for more recess and more roughhousing.)

maggiekb (Maggie Koerth-Baker, senior science writer): I’ve wondered how much of this is about how you distinguish yourself, as a guy, from the other candidate when that other candidate is a woman. Whether the hyper-masculinity is growing out of a place of knowing, essentially, that you’re running against Hillary Clinton.

natesilver: Yeah, this is hard to prove, but my sense is that the chest-pounding / dick-measuring part of the GOP race doesn’t have all that much to do with Hillary Clinton.

simone: Why not?

natesilver: For one thing, because she’s not the president yet — or even officially the Democratic nominee yet — and the outgoing president typically has a much bigger role on the other party’s psyche than the incoming one. I mean, the racism has also been a lot more explicit on the GOP side, at least when coming from Trump. So maybe we need — dare I say it — an intersectional theory to explain it.

clare.malone: See my above “traditional culture” rant! Republicans see the Democratic Party and the popular culture at large moving farther and farther away from what they’re used to, including how men are viewed — that’s why things like transgender bathrooms have become such an issue, for instance.

harry: I think it has more to do with society overall. A ridiculously high 68 percent of Trump supporters say society is becoming too soft and feminine. Cruz and Kasich supporters come in with 57 percent and 52 percent, respectively. Now compare those numbers with the Democratic side, where Sanders supporters were slightly less likely than Clinton supporters to say that (28 percent vs. 31 percent).

maggiekb: Interestingly, Nate, there’s been some academic literature describing Obama as a unisex candidate — that part of what worked for him was the ability to hit both ends of the stereotype spectrum when he needed to.

simone: Then how do we explain the “Bernie Bro” phenomenon on the Democratic side? Sanders is quite literally running against Clinton, and his supporters have been accused of sexism. But is anyone accusing Sanders of being masculine?

leah: Well, there’s a kind of masculinity he does fit into (the cranky old uncle model). And those behaviors (the loudness, the aggressiveness) play very differently for a man than a woman. There’s a lot that’s treated as adorable from him that wouldn’t pass from Clinton.

clare.malone: Trump would likely try to paint Sanders as just a batty old radical. I don’t think he could do an emasculation spin on Sanders.

natesilver: Behavior changes whenever you get groups of men together. Especially young men, and especially on the internet.

harry: I don’t particularly think of the Bernie Bro as masculine.

simone: You think they’re “betas,” harry?

harry: Here’s the No. 1 image for Bernie Bro …

clare.malone: Beta males can be very masculine! I tend to think that alpha/beta split refers more to competitive drive and industry than masculinity — e.g., Aidan from “Sex and the City” — beta. Yeah, I said it. Mr. Big = alpha.

maggiekb: But he’s also masculine in that sweaty, building-stuff way — Aidan is. And Mr. Big is feminine in the sense of tailoring, fine dining, haircuts.

harry: John on “Sex and the City” was quite masculine. Aidan was a little punk whose best role was in “My Big Fat Greek Wedding.”

simone: This is digressing. OK, so some of the candidates are flirting with hyper-masculinity. Do we have any evidence that voters care? Are men (or women) motivated to go to the polls by manliness?

leah: It’s hard to say — there aren’t any studies I’ve found that look at how persuasive direct allusions to genitalia are! Who would have thought we needed those numbers.

maggiekb: There is pretty good evidence that, depending on the year, perceived masculinity can matter a lot to voters. But the catch is that bit about the year. Because we attach gender stereotypes to issues.

natesilver: I’ll remind people here that Donald Trump is one of the most unpopular politicians in the history of the American republic.

harry: Thanks, Nate. I can say there is some evidence that women voters on the Democratic side are more likely to vote for Clinton when they believe they have been personally discriminated against.

Listen to the latest episode of the FiveThirtyEight elections podcast.

http://c.espnradio.com/s:5L8r1/audio/2725535/fivethirtyeightelections_2016-04-11-183435.64k.mp3?ad_params=zones%3DPreroll%2CPreroll2%2CMidroll%2CMidroll2%2CMidroll3%2CMidroll4%2CMidroll5%2CMidroll6%2CPostroll%2CPostroll2%7Cstation_id%3D4278Subscribe: iTunes |Download |RSS |Video

maggiekb: So here’s a question: Is Trump masculine? We’re coming at this from the assumption that he represents aggressive masculinity, but is that really true? Are voters perceiving Trump as masculine?

clare.malone: I think they are — part of the way that he’s escaped the coddled rich boy mold is by being a blustering male type, talking about his sexual exploits, talking about his risk-taking — he’s unpretentious in that way. He reads as a guy in the locker room.

natesilver: They perceive Trump as authentic and that Trump uses displays of masculinity to assert his authenticity.

simone: And setting himself up in opposition to “Lil Marco” Rubio.

natesilver: Rubio is also an interesting figure in the GOP race, in my view.

simone: Was, Nate, was.

leah: He hasn’t released his delegates!

clare.malone: In, but just on a technicality (that’s what she said).

harry: I don’t see Trump as masculine, but that’s because my view of masculinity doesn’t mean “macho.” Rubio, on the other hand, along with Jeb Bush (to a greater extent), wanted to talk about issues. Rubio found that wasn’t working and went the Trump route. He found that perhaps there was more behind Trump’s support than dirty language.

natesilver: But remember that Rubio’s boyish looks worked against him. Rubio is actually about the same age as Cruz, but voters were much more likely to say that he was too young to be president. Republicans were also making a lot of implicit comparisons between Rubio and Obama, however, another young-looking guy.

simone: And non-white guy.

natesilver: Wait, now I have a theory about Bernie, though. The fact that he’s a little rumpled, rough-around-the-edges — is that also a reaction to Obama in some way?

clare.malone: As the more authentic revolutionary, Nate? Yeah, maybe.

leah: Again, it’s hard to pin down what masculinity is in order to research it. Boyish looks are actually one of the things we do have numbers on, and the numbers aren’t good for Rubio or men who look like him. Researchers have shown people pictures of candidates running in gubernatorial and U.S. Senate races and asked them to pick the one who looked more competent after seeing both for a split second. The one perceived as more competent tended to win way more often than chance. And “competent” tends to equal conventionally masculine features (strong jaw, etc.), not Rubio’s younger looking, rounder face. Just looking at that perceived competence predicted 69 percent of gubernatorial races and 72 percent of senatorial ones, according to a 2007 paper.

clare.malone: We’ve all been conditioned in certain ways — and even if you’re down with feminism and up on the correct way to address changing gender norms, if you were born in a certain year (let’s say pre-1995), you have some pretty ingrained bias you might never get rid of. Those Snapchatting younger millennials, now, they might be a different story when it comes to perceptions of candidates in elections to come over the next decades.

harry: Women make up more than 50 percent of the electorate in the general election. If Trump pulls any of the stuff he pulled in the primary, it will come across far different. The reason? Because of the way we view gender roles.

clare.malone: Trump, if he’s the nominee against Clinton, is going to inspire a lot of Democratic women to get out there and vote just on the basis of the Howard Stern transcripts alone.

maggiekb: He might inspire some Republican women to get out and vote … for Clinton

clare.malone: Or stay home, which ain’t great either.

simone: But will Trump inspire men to get out there and vote? I guess what I’m trying to get at is whether we can also talk about men as an identity/interest group in the way we often talk about women and racial and ethnic minorities. Seems like you guys (and gals) are saying the answer is no, or at least as we’ve seen so far in voter behavior? And maybe that tells us something about the supposed cohesiveness of those other supposed identity groups.

maggiekb: That’s a really interesting point, Simone. I was thinking about that after looking at some surveys of voter interest in certain issues. Like this one:

We talk about things like education as being a woman’s issue. We don’t talk about taxes as being a man’s issue.

clare.malone: I’ve talked to a lot of Republican women who say they don’t like to be pigeonholed as “women voters.” They’re concerned with the corruption of the political establishment, or whatever their particular issue of concern may be — there’s not a great perception of “feminism” in a lot of corners of the GOP, because there’s a notion that it makes women one-dimensional members of the electorate. Which is, of course, not the case on either side of the aisle — women, like men, contain multitudes when it comes to motivations and their politics.

leah: Plus, abortion access, which is traditionally treated as a get-out-the-vote issue for women, is actually polarizing — there are a lot of pro-life women voters.

maggiekb: If there’s one thing that’s moving us more toward men as a special interest group, rather than the default, it is the focus, in this election, of framing Trump’s supporters in the white, blue-collar, angry man category. So maybe that’s the new soccer mom?

natesilver: There’s definitely a sense in which gender blends into other types of political identity — more so than race does, frankly. To put this in somewhat abstract terms, if you run a big logistic regression analysis with a bunch of characteristics that predict someone’s vote, the coefficient for gender doesn’t usually show up as being all that significant.

harry: Well, if we look back at the exit polls in the primary contests, we already see some backlash against Trump with women. In some states, Trump gets about an equal percentage of the male and female vote (e.g., Wisconsin). But in many states, there’s a large difference. South Carolina (7 percentage points), Alabama (16 percentage points), Arkansas (6 percentage points), Georgia (10 percentage points), Massachusetts (6 percentage points), Oklahoma (9 percentage points). Shall I go on?

clare.malone: moar polls.

simone: How does that bode for the general?

harry: These splits are roughly equivalent to Romney’s split between male and female voters in the 2012 general election. If Trump is already doing relatively poorly among Republican women, then how well can he do when he faces non-Republican women who tend to be more Democratic? Indeed, Trump trails Clinton by 12 percentage points among married women in a new poll. Romney won married women by 7 percentage points in 2012.

maggiekb: I feel like trying to talk about the “male vote” and “male voters” here really exposes some of the inherent flaws in talking about “female voters” and the “female vote” as one block, per leah.

simone: So does Clinton’s position in the race (and let’s not forget Carly Fiorina on the GOP side) somehow change the game going forward? Whether she wins the nomination or even the presidency, are we in for a big shift in how we think about candidates and gender?

leah: Well, now I get to talk about my all-time favorite natural experiment. In India, different villages were randomly assigned to be run by women (only women were allowed to be candidates in the elections), and that meant researchers could test how much being governed by a woman can shake up people’s preferences for male and masculine leaders.

In places that were forced to have women leaders, folks still preferred men explicitly, but their ratings of how competent women were went up. And, once they’d been forced to have women twice, they were significantly more likely to elect women in open elections.

So, after a couple of women presidents, we might trust women more. Yay?

clare.malone: Oy.

maggiekb: I love this study, Leah! I mean, I don’t love the outcome. But it’s fascinating.

I was also really struck, in trying to research this, how much of the science of gender and elections is about femininity/women and elections. We can have a conversation about the science of women and politics and never really talk about men or male candidates much, but we can’t do that the other way around. Male/masculinity is in some ways the assumed default, and we’re doing research on what happens when we diverge from that norm, but there’s not a lot of research on what the norm means. Whether the norm is changing. How the norm is adapting itself when it’s forced to run against something other than a male. I really want to see that research going forward.

simone: Does that mean we get to stop talking about dicks in 2020?

clare.malone:

Elections Podcast: Is Trump Blowing It?

Our elections podcast team takes stock of the recent Republican state conventions where Donald Trump failed to organize his supporters and secure delegates. Plus, in the first installment of Harry’s History, a lesson in strategic voting from New York’s 1992 Democratic primary. Also, in case you missed our FAQ, find it here or in iTunes.

http://c.espnradio.com/s:5L8r1/audio/2725535/fivethirtyeightelections_2016-04-11-183435.64k.mp3?ad_params=zones%3DPreroll%2CPreroll2%2CMidroll%2CMidroll2%2CMidroll3%2CMidroll4%2CMidroll5%2CMidroll6%2CPostroll%2CPostroll2%7Cstation_id%3D4278Subscribe: iTunes |Download |RSS |Video

You can stream or download the full episode above. You can also find us by searching “fivethirtyeight” in your favorite podcast app, or subscribe using the RSS feed. Check out all our other shows.

If you’re a fan of the elections podcast, leave us a rating and review on iTunes, which helps other people discover the show. Have a comment, want to suggest something for “good polling vs. bad polling” or want to ask a question? Get in touch by email, on Twitter or in the comments.

Trump’s New Magic Number Is 40 Percent Of The Vote

Donald Trump has 39 percent of the vote in our Pennsylvania polling average, 37 percent in California, and 39 percent in Maryland. If this were February or early March, that would leave him without much to worry about. Even if Trump picked up zero undecided voters, he’d be pretty much guaranteed a win with the rest of the vote divided between a half-dozen opponents.

But those days are over. In Wisconsin on Tuesday, Trump had 35 percent of the vote — the same share that allowed him to win New Hampshire easily in February, and a larger percentage than he got in winning South Carolina. But not only did Trump fail to win Wisconsin — he got crushed by Ted Cruz.

In many respects, this is an old story. One of the main reasons for our initial skepticism of Trump’s candidacy last summer and fall is that his high unfavorable ratings implied he’d have trouble gaining ground as the field winnowed, potentially allowing other candidates to overtake him.

But I think people may under-appreciate the degree to which this is no longer just a theoretical problem for Trump. It’s become an actual problem, as is readily apparent in the data from the states that have voted so far. The threshold Trump needs to win states is increasing considerably faster than the share of the vote he’s getting, which isn’t increasing much at all. Technically, Trump is chasing delegates, not wins, but most of the remaining states award at least some delegates to the statewide winner (and there are still five winner-take-all contests left on the GOP calendar).

So that we can be more precise about this, I’m going to define a statistic called the Minimum Winning Vote Share. As its name implies, it shows the smallest percentage of the vote a candidate could receive and still win a state. Here’s how we’d calculate it for Trump in South Carolina, for instance. We start by listing the number of votes received by every candidate except Trump:

CANDIDATEVOTESRubio166,565Cruz165,417Bush58,056Kasich56,410Carson53,551South Carolina vote totals, excluding Trump

Source: The Green Papers

Marco Rubio was the top non-Trump candidate, receiving 166,565 votes. In order to win, Trump needed one more vote than Rubio, or 166,566 votes. If Trump had gotten that many votes, with all other candidates staying the same, he’d have won South Carolina with only 25 percent of the vote. That’s Trump’s Minimum Winning Vote Share.

CANDIDATEVOTESSHARETrump166,56625.0%Rubio166,56525.0Cruz165,41724.8Bush58,0568.7Kasich56,4108.5Carson53,5518.0Hypothetical South Carolina results, with Trump winning by one vote

In actuality, Trump got 240,882 votes in South Carolina, winning the state fairly easily. But those extra votes were superfluous; a quarter of the vote would have gotten it done. Trump’s Minimum Winning Vote Share is typically much higher than that now, however. Here it is for all states (plus the District of Columbia) to have voted so far:

TRUMP VOTE SHAREDATESTATEMINIMUM REQUIRED TO WINACTUAL2/1Iowa26.7%24.3%2/9New Hampshire19.535.22/20South Carolina25.032.52/23Nevada30.645.93/1... Carolina38.140.2Missouri40.840.9Ohio4... minimum winning vote share

Trump’s lowest Minimum Winning Vote Share was in New Hampshire, where he could have gotten away with just 19.5 percent of the vote and still beat John Kasich.1 On Super Tuesday, Trump’s average Minimum Winning Vote Share was just 31.2 percent.

But as I said, it’s been increasing steadily. It was 37.4 percent on average in the five states to vote on March 15. And it’s averaged 40.3 percent in the three states to vote since then, including 42.6 percent in Wisconsin. Here’s the same data in graph form:

While there will continue to be some variance from state to state, Trump is now usually going to have to be in the 40s to win. That’s a problem, because as you can see from the bottom half of the chart, it’s not clear that his performance is improving much at all. (This is also apparent in national polls, where Trump’s share of the vote has grown only to 40 percent from 35 percent before Iowa.) Of the six states where Trump’s Minimum Winning Vote Share has been at least 40 percent – Maine, Mississippi, Missouri, Ohio, Utah and Wisconsin – Trump has won only two.

Part of the problem is that Republican voters seem to be behaving tactically, gravitating toward the most viable non-Trump alternative in their state. In Wisconsin on Tuesday, Cruz beat his polling average by about 10 percentage points. But Kasich underperformed his by 5 points, suggesting the presence of a #NeverTrump vote.

Furthermore, because Trump tends to do poorly with late-deciding voters and doesn’t have much of a turnout operation, Trump has tended to hit his polling averages right on the nose instead of gaining from undecided voters. If he’s at 37 percent in the polling average in a state, that’s a reasonably good estimate of his election-day vote; you don’t necessarily want to round up a few points to account for undecideds, as you would for most candidates.

Listen to the latest episode of the FiveThirtyEight elections podcast.

http://c.espnradio.com/s:5L8r1/audio/2720801/fivethirtyeightelections_2016-04-05-163601.64k.mp3?ad_params=zones%3DPreroll%2CPreroll2%2CMidroll%2CMidroll2%2CMidroll3%2CMidroll4%2CMidroll5%2CMidroll6%2CPostroll%2CPostroll2%7Cstation_id%3D4278Subscribe: iTunes |Download |RSS |Video

Trump will probably still be fine in the Northeastern states to vote later this month. In New York, Trump’s above 50 percent in polls, so he doesn’t care how Kasich and Cruz split up the rest of the vote. And in Pennsylvania and Maryland, Cruz and Kasich have competing claims for being the best anti-Trump, which could complicate tactical voting.

Once we leave the Northeast, however, Cruz is liable to be much more viable than Kasich, allowing him to gain from tactical voting. I’d be nervous about California if I were Trump, for example – Cruz is only 6 or 7 percentage points behind, there are two months to go, and 33 percent of voters are either undecided or say they’ll vote Kasich. If past states are any guide, a pretty decent fraction of those votes could wind up with Cruz.

It may still wind up being too little and too late for Trump’s opponents, who would have benefited from a smaller field early on. But Trump will have to work harder to win states the rest of the way out.

Ted Cruz, Not Paul Ryan, Would Probably Win A Contested Convention

It’s like something out of an Aaron Sorkin script. After their bitterly divisive primary, the Republican delegates come together to nominate John Kasich on the fourth ballot at a contested convention in Cleveland, despite his having won only his home state of Ohio. Or they choose House Speaker Paul Ryan, despite his not having run in the primaries at all. Balloons descend from the ceiling, celestial choirs sing and everything is right again with the Republican Party, which goes on to beat Hillary Clinton in a landslide in November.

As I said, it’s like something out of a TV show. In other words: probably fiction. It’s not that hard to imagine a contested convention. In fact, with Donald Trump’s path to 1,237 delegates looking tenuous, especially after his loss in Wisconsin on Tuesday night, it’s a real possibility. And it’s not hard to see how Republicans might think of Kasich or Ryan as good nominees. If Republicans were starting from scratch, both might be pretty good picks, especially from the perspective of the party “establishment” in Washington.

But Republicans won’t be starting from scratch, and the “establishment” won’t pick the party’s nominee. The 2,472 delegates in Cleveland will. And most of them will be chosen at state or local party conventions a long way from Washington. Few will be household names, having quietly attended party gatherings in Fargo, North Dakota, or Cheyenne, Wyoming, for years with little remuneration or recognition. Although the proverbial Acela-riding insiders might dream of Ryan or Kasich, there are indications that the rank-and-file delegates are into Ted Cruz — and they’re the ones who will have votes in Cleveland.

To recap a bit, the Republican presidential voting process is separate from the delegate selection process in most states. In South Carolina, for instance, most delegates are selected through a series of county, congressional district and state conventions. Although those delegates are bound to Trump (who won the state’s primary on Feb. 20) on the first ballot, they could peel off and vote for another candidate after that.1

There are some states where delegates are selected directly on the ballot (as in Maryland, for instance) and others where slates are submitted by the candidates (as in New Hampshire) — these are a fairly small minority. Below, you’ll find a table showing the Republicans’ delegate selection method in all states and territories, according to the Republican Party’s rulebook.

TIED TO PRES. PREF. VOTENOT TIED TO PRES. PREF. VOTESTATE OR TERRITORYBY CANDIDATESON PRIMARY BALLOTAT STATE OR LOCAL CONVENTIONSBY STATE OR LOCAL PARTY COMMITTEESRNC MEMBERSAlabama473Alaska253American Samoa63Arizona553Arkansas12253California1693Colorado343Connecticut253Delaware133D.C.163Florida81153Georgia733Guam63Hawaii163Idaho293Illinois54123Indiana543Iowa273Kansas373Kentucky433Louisiana433Maine203Maryland24113Massachusetts27123Michigan563Minnesota353Mississippi373Missouri493Montana243Nebraska333Nevada273New Hampshire203New Jersey483New Mexico213New York923North Carolina693North Dakota253No. Mariana Isl.63Ohio633Oklahoma403Oregon253Pennsylvania54*143Puerto Rico203Rhode Island163South Carolina473South Dakota263Tennessee41143Texas1523Utah373Vermont133Virgin Islands63Virginia463Washington413West Virginia313Wisconsin24153Wyoming263Total2593981,358289168Share of total10%16%55%12%7%How are Republican delegates chosen?

* 54 Pennsylvania delegates are directly elected but unbound to any candidate.

SOURCE: Republican National Committee Presidential Process Planning Book

Without getting too lost in the details,2 there are five major delegate selection methods:

Candidates choose their delegates (10 percent of delegates). In some states, candidates name a slate of delegates. These states include California, making it even more important to the Republicans’ delegate math; delegates won in California are likely to remain loyal to their candidates longer than in most places.Directly elected (16 percent of delegates). Other delegates, as I mentioned, are chosen directly on the primary ballot. Usually, the ballot indicates which candidate the delegate prefers, and the delegates are bound to that candidate. An important exception is Pennsylvania, where 54 delegates will be elected on the ballot as uncommitted.

In these first two cases, there’s a strong link between the presidential preference vote and delegate selection. The link isn’t perfect —

Momentum May Matter (Just This Once) In Wisconsin

For this week’s politics Slack chat, we unpack what’s riding on today’s Republican and Democratic primaries in Wisconsin. The transcript below has been lightly edited.

Check out our live coverage of the Wisconsin primary elections.

micah (Micah Cohen, politics editor): After a slow couple of weeks, it’s election day again! Wisconsin, home of cheese and badgers, holds both a Democratic and Republican primary today. So let’s preview both races, starting with the GOP: Harry, what do we expect to happen tonight on the Republican side?

clare.malone (Clare Malone, senior political writer): Now I have images in my head of badgers just rolling around in cheese curds.

harry (Harry Enten, senior political writer): Well, if the polls are right (minus that weird American Research Group survey), Ted Cruz should roll into Wisconsin and eat most of if not all the cheese. Cruz either has an 80 percent (polls-only) or 89 percent (polls-plus) chance of winning, depending on which FiveThirtyEight model you look at. The real question, though, is how many congressional districts he wins. If Cruz sweeps statewide and all the congressional districts, he takes all of Wisconsin’s 42 delegates. Donald Trump, however, has a shot in at least two districts (the 3rd and the 7th) in the north and west part of the state. That means we could be looking at a delegate split along the lines of Cruz 36/Trump 6/John Kasich 0, or Cruz 33/Trump 3/Kasich 3 (if Kasich wins around Madison in the 2nd District).

natesilver (Nate Silver, editor in chief): Indeed, Cruz could poke holes in Trump’s quest for 1,237 delegates like he’s a slice of Swiss cheese.

micah: Clare, why isn’t Trump doing better in Wisconsin? Did we always expect this to be a tough state for The Donald?

clare.malone: Well, I think there’s something to be said for the fact that Scott Walker and Paul Ryan are all up in Wisconsin and very much not Trumpian Republicans. They’re very likely operating behind the scenes, getting the local party apparatus working against Trump.

micah: And the GOP apparatus in Wisconsin is strong, no?

clare.malone: Right, Wisconsin typically sees higher turnout than most states in most elections — Wisconsinites are very politically engaged!

natesilver: I suppose I find the Wisconsin-is-just-a-bad-state-for-Trump a little bit circular.

harry: Please tell us more, Nathaniel.

natesilver: When our expert panel convened to game out the rest of the states a few weeks ago, we had Trump winning the majority of delegates in Wisconsin. And I’d guess that was pretty in line with the conventional wisdom at the time. Trump won Michigan fairly easily, which is a relatively similar state to Wisconsin. He won Illinois. He lost Iowa, but it was a caucus — Wisconsin holds a primary, and those have tended to favor Trump.

micah: So you don’t buy this article from friend-of-the-site Nate Cohn at The New York Times, which argued that Trump’s “problem in Wisconsin is mainly about the state’s demographics, not self-inflicted wounds. Even a 10-percentage-point loss there wouldn’t necessarily indicate any shift against him.”

natesilver: I’d buy a diet version of that argument. I think it’s a bit overstated. The fact is that the Republican primary is pretty hard to model demographically.

clare.malone: I love that there’s a theory out there that Wisconsinites are more “Yankee” than other Midwestern states, somehow. Elections really bring out our trafficking in stereotypes, don’t they?

natesilver: As we talked about on the podcast Monday, Wisconsinites (and other people in the upper Midwest) have high levels of social connectivity, which seems to be a correlate of the #NeverTrump vote. Trump voters are fairly socially isolated, by contrast.

clare.malone: I actually really am intrigued by that theory. It also ties in with the feeling of certain Trump voters in states with higher black populations feeling like persecuted whites in an era of increased attention to racial sensitivities and inequities.

Wisconsin’s pretty white, so those Yankee/Scandinavian descendants probably aren’t feeling as embattled.

micah: But there are dozens and dozens of demographic variables we could look at to explain Trump’s support, including social connectivity, so some are bound to correlate pretty well with how he’s done. Should we really be putting that much stock in these demographic-based findings on the GOP side (including our own)?

natesilver: Yeah, I definitely think there’s a lot of overfitting going on in people’s mental (and mathematical) models of Trump. Which is why, on some level, a loss would be a loss tonight and shouldn’t be totally explained away. Also, Wisconsin is one of the first examples we’ll have gotten of what happens to the vote now that Rubio’s dropped out. In Arizona, we didn’t get a clean test because so much of the vote was cast early with Rubio still running. And Utah — now there’s a place where you can excuse Trump’s performance based on demographics because they’re so homogeneously Mormon. So Wisconsin is the best test we have so far of what a three-way race looks like.

harry: Key point there from the man from Michigan — I think the margin and Trump’s percentage of the vote are as important as whether Cruz wins.

If Trump hits, say, only 33 percent, it’ll show that he picked up pretty much no ground after Rubio dropped out. I base that off of Marquette polls taken before and after Rubio exited. If however, Trump is in the high 30s, it’ll signal to me that he did indeed pick up some of that Rubio support.

Listen to the latest episode of the FiveThirtyEight elections podcast.

http://c.espnradio.com/s:5L8r1/audio/2720801/fivethirtyeightelections_2016-04-05-163601.64k.mp3?ad_params=zones%3DPreroll%2CPreroll2%2CMidroll%2CMidroll2%2CMidroll3%2CMidroll4%2CMidroll5%2CMidroll6%2CPostroll%2CPostroll2%7Cstation_id%3D4278Subscribe: iTunes |Download |RSS |Video

micah: So this gets to another question: I’ve heard a lot of talk that a Cruz win in Wisconsin could “reset the race” — do we buy that?

natesilver: It’s important to remember that “momentum,” as the press talks about it, is backward-looking rather than forward-looking. There’s a perception that Trump’s had a bad couple of weeks — he’s now below 50 percent to win the nomination at prediction markets, which seems awfully cheap — and a loss in Wisconsin would confirm that perception.

clare.malone: I don’t think I buy the wholesale “reset” thing anymore, because I think we’re basically just headed into a period of extended litigation on the ground with party meetings and electing delegates and obscure rules and all that. I tend to think that’s the direction of the race now. I think the way a Wisconsin win might help things for the Cruz team and hurt Trump is that it helps in the Cruz campaign’s delegate-courting process, their convincing people that he’s viable.

natesilver: Yeah, so that’s one way that momentum (real or perceived) could matter. While there aren’t a lot of states voting over the next couple of weeks, there are a lot of them picking delegates. And although Cruz is doing well in that delegate-selection process already, it’s an easier sell when he can say the wheels are coming off Trump’s campaign. By contrast, if Trump wins Wisconsin — unlikely, but stranger things have happened several times this year already — the Cruz campaign’s delegate-related machinations could look more like a naked power grab. And the conventional wisdom, which could matter to the GOP delegates, could shift back to “how can we NOT give Trump the nomination given that he clearly has the most popular support?” instead of “how can we possibly nominate this guy?”

clare.malone: “Naked Power Grab” is the sequel to my romance novel “Delegates Unbound.” The Trump having a bad delegate-hunting-machine line is actually going to be really interesting, I think. And how litigious things might get — he’s certainly rubbed a lot of state GOP’s the wrong way over the last week to 10 days.

harry: Well, let’s put it this way: If Cruz can win Wisconsin by a wide margin, sweep the delegate count and then something happens to keep Trump under 50 percent in New York, then things will get very interesting. If, however, Trump merely sets a Wisconsin loss aside and then wins in New York and the rest of the states voting this month, I’m not sure winning Wisconsin by 5 percentage points versus 10 points versus 15 matters in the grand scheme. Of course, a large Wisconsin win for Cruz probably means he’ll go on to win the winner-take-all states of Nebraska and South Dakota.

natesilver: It’s true that other things held equal, Trump is likely to be underrated after Wisconsin (assuming he loses) and overrated after New York (assuming he wins big). That’s not the same thing as saying momentum doesn’t matter, however. Perception could matter quite a lot to the delegates, and they’re the ones who might end up deciding the nomination.

harry: I guess I don’t know what the perception would actually be in the scenario above.

clare.malone: Well, in Harry’s scenario, the perception playing field would be more even, right?

natesilver: Yeah. If Trump wins out every time he has a big confrontation with Cruz (and Kasich), it’ll look more like he has a mandate from GOP voters. And that will make it harder for delegates to deny him the nomination. Cruz needs Wisconsin precisely because he’s going to take some losses later in the month. Not just for the delegates — we’re already “pricing in” Trump doing very well in New York, Connecticut, etc. — but also for the morale of the #NeverTrump forces and for the argument Cruz will make to the delegates in Cleveland.

harry: I’m left wondering more about how big these wins are for each of them.

natesilver: But as much as we might say “it’s all about the math,” Harry, the fact is that the delegates might have to make this very subjective decision about whether nominating Trump or denying him the nomination is the least-worst outcome for the GOP. It’s all about the math if Trump gets 1,237 delegates — well, unless the delegates want to pull a fast one and unbind themselves — but in the zone where he’s somewhere around 1,150 or 1,200 delegates, the narrative the delegates tell themselves about the race matters a lot.

harry: Right, and I guess I’m saying I don’t know how the delegates will react. And I’d think the margins may matter in that case. Or maybe, they won’t. I have no clue.

natesilver: How the media covers the race matters a lot too. They’ve finally gotten away from the Trump-is-invincible storyline. If he pulls it out in Wisconsin, however, they’ll go right back to it.

clare.malone: We’ve reached a very sticky wicket wherein we’re now dabbling in individual’s reactions to the state of the country rather than demographics. If Trump’s bad couple weeks continue, and stories about, say, campaign “disarray” and personnel problems keep up, there might be some on-the-fence delegates who won’t swing toward Trump after all.

natesilver: Yeah, this just seems like a poorly chosen time for the “momentum doesn’t matter” line. I agree with that line 90 percent of the time. This case might be in the 10 percent for the GOP.

clare.malone: That’s a pretty atypical Nate Silver statement! But it’s an atypical time.

micah: Yeah, mark down this moment “Nate Silver Endorses The Big Mo’!”

harry: Nate has gone mainstream media. Sad day for journalism.

natesilver: Just for the Republicans, though! I don’t think momentum matters that much for the Democrats.

clare.malone: Democrats … what of them, eh? I feel like we’re about to give our standard lines about the race, which is that math don’t lie and that while Sanders will likely win Wisconsin, it might not mean all that much. Go.

harry: The Democratic side is fairly simple. Sanders has led in most of the polls of Wisconsin, and the state’s demographics suggest a mid-single-digit win for him. Given Clinton’s 224 pledged delegate lead, Sanders probably needs to win by about 16 percentage points or more to be on track to come all the way back and win the nomination.

micah: So let’s say Sanders wins Wisconsin by 15-plus percentage points — what would that mean for the race overall?

harry: It would mean that maybe something has changed. Maybe Sanders has a real shot at winning this thing. Of course, it doesn’t mean anything if Clinton goes on to win New York by double-digits.

natesilver: I’m a bit surprised that the Wisconsin polls have it so close. Then again, we’ve been conditioned to think in terms of caucuses when we’re back to primaries now.

clare.malone: How much do we trust Wisconsin polls?

natesilver: The polls have been pretty crappy overall on the Democratic side, but they’ve missed in different directions in different states.

clare.malone: So, with Sanders up 3 percentage points in our weighted average of Wisconsin polls, is a win of 15-plus points unrealistic?

natesilver: The average poll in the Democratic race has been something like 10 points off. Polling averages do a fair bit better

harry: Within 21 days of the caucus or primary, the average poll has been off 11 percentage points so far this campaign. But as Nate said, it’s been a mixed bag. In Ohio, which everyone thought could be a Sanders state after he won Michigan, the polls were dead-on.

micah: And the polling error hasn’t consistently been in either Clinton or Sanders’s direction?

natesilver: Bernie winning by 15 would be a big deal, but not a huge, epic surprise. Like a magnitude 6.0 earthquake, not an 8.0.

harry: Right, it’d put me into a more of a wait-and-see mode.

natesilver: The polls have tended to underrate Clinton in the South and underrate Bernie in the North. Which would be a good sign for Sanders in Wisconsin. I suspect the polls are having a lot of trouble picking up enough black voters and young voters, which are the two most decisive groups in the Democratic primary.

With all that said, the contrarian devil on my shoulder is saying that since it seems obvious to everyone that Sanders will beat his polls in Wisconsin, maybe we’re heading for a Clinton upset instead.

micah: So let’s end on that: If Clinton does win in Wisconsin, is this race over? I mean, it’s close to over now, according to the delegate math, but would a Bernie loss in Wisconsin lead to an “R.I.P.” narrative too?

clare.malone: The race won’t be over until the very last state — technically speaking — and because no matter what happens (Clinton winning in Wisconsin), Sanders probably won’t drop out.

natesilver: As Harry said, New York is the linchpin state on the Democratic side. Both because it has a lot more delegates than Wisconsin and because it might give us an indication of how the April 26 states — and to a lesser extent, California on June 7 — will vote.

micah: So even if Clinton wins Wisconsin, we have to wait until New York? Or, to Clare’s point, June 7? Sanders supporters seem unlikely to concede no matter what the results.

clare.malone: It’s gonna be a subway series.

Elections Podcast: The Delegate Hunt Begins

We’re coming to D.C! We just announced our first-ever live show, in Washington on May 4. Get your tickets now, and help spread the word to anyone in the D.C. area. (And don’t worry, we’ll be doing more live shows in other cities throughout the summer and fall.)

In this episode, the elections podcast crew begins to track the Republican candidates’ hunt for sympathetic delegates as the possibility of a contested convention appears to grow. Then, producer Galen Druke discusses the ins and outs of Wisconsin’s political landscape in the run-up to the state’s primary on Tuesday. Plus: How would you describe Donald Trump’s coalition to someone who hasn’t been following the election?

http://c.espnradio.com/s:5L8r1/audio/2719954/fivethirtyeightelections_2016-04-04-182054.64k.mp3?ad_params=zones%3DPreroll%2CPreroll2%2CMidroll%2CMidroll2%2CMidroll3%2CMidroll4%2CMidroll5%2CMidroll6%2CPostroll%2CPostroll2%7Cstation_id%3D4278Subscribe: iTunes |Download |RSS |Video

You can stream or download the full episode above. You can also find us by searching “fivethirtyeight” in your favorite podcast app, or subscribe using the RSS feed. Check out all our other shows.

If you’re a fan of the elections podcast, leave us a rating and review on iTunes, which helps other people discover the show. Have a comment, want to suggest something for “good polling vs. bad polling” or want to ask a question? Get in touch by email, on Twitter or in the comments.

It’s Probably First Ballot Or Bust For Donald Trump At The GOP Convention

At the prediction market Betfair on Friday morning, bettors put Donald Trump’s chances of winning the Republican presidential nomination at 56 percent. That’s down a fair bit — Trump had been hovering at about 70 percent after his win in Arizona (and loss in Utah) last week. Meanwhile, the likelihood of a contested convention according to bettors has considerably increased. There’s now a 63 percent chance1 that the convention in Cleveland will require multiple ballots, according to Betfair.

In other words, the markets are now betting on a contested convention. Not just a near-miss, where the nomination is resolved at some point between the last day of GOP primaries June 7 and the start of the convention July 18, but the thing that political journalists dream about: a full-blown contested convention where it takes multiple ballots to determine the Republican nominee.

Here’s the thing, though: Those markets don’t make a lot of sense. If you really think the chance of a multi-ballot convention is 63 percent, but also still have Trump with a 56 percent chance of winning the nomination, that implies there’s a fairly good chance that Trump will win if voting goes beyond the first ballot. That’s probably wrong. If Trump doesn’t win on the first ballot, he’s probably screwed.

The basic reason is simple. Most of the 2,472 delegates with a vote in Cleveland probably aren’t going to like Trump.

Let’s back up a bit. In most of our discussions about delegates here at FiveThirtyEight, we treat them as though they’re some sort of statistical unit. We might say a candidate “racked up 44 delegates” in the same way we’d say Steph Curry scored 44 points. But those delegates aren’t just a scoring mechanism: Delegates are people, my friends. Delegates are people!

And as I said, they’re mostly people who aren’t going to like Trump, at least if the excellent reporting from Politico and other news organizations is right. (If Trump turns out to have more support among GOP delegates than this reporting suggests, even marginally, that could end up mattering a great deal.) How can that be? In most states, the process to select the men and women who will serve as delegates is separate from presidential balloting. In Massachusetts, for instance, Trump won 49 percent of the GOP vote on March 1 — his highest share in any state to date — to earn 22 of the state’s 42 delegates. But the people who will serve as delegates haven’t been chosen yet. That will happen at a series of congressional district conventions later this month and then a Republican state meeting in May or June. According to Politico, most of those delegates are liable to favor Ted Cruz or John Kasich rather than Trump. Twenty-two of them will still be bound to Trump on the first ballot, but they can switch after that. The same story holds in a lot of other states: in Georgia, Louisiana and South Carolina, for instance — also states that Trump won.

Trump’s delegate problems stem from two major issues. One is his lack of organization: Trump just recently hired a strategist to oversee his delegate-selection efforts; Cruz has been working on the process for months. The other is his lack of support from “party elites.” The people who attend state caucuses and conventions are mostly dyed-in-the-wool Republican regulars and insiders, a group that is vigorously opposed to Trump. Furthermore, some delegate slots are automatically given to party leaders and elected officials, another group that strongly opposes Trump, as evident in his lack of endorsements among them.

There are various ways these delegates could cause problems for Trump. The most obvious, as I mentioned, is if the convention goes to a second ballot because no candidate wins a majority on the first. Not all delegates become free instantaneously,2 but most do, and left to vote their personal preference, most of them will probably oppose Trump.

Conversely, Trump isn’t totally safe even if he locks up 1,237 delegates by the time the final Republicans vote. The delegates have a lot of power, both on the convention floor and in the various rules and credentials committees that will begin meeting before the convention officially begins. If they wanted to, the delegates could deploy a “nuclear option” on Trump and vote to unbind themselves on the first ballot, a strategy Ted Kennedy unsuccessfully pursued against Jimmy Carter in 1980.

Although I’d place fairly long odds against this thermonuclear tactic, there’s also the possibility of piecemeal skirmishes for delegates. In South Carolina, for instance, delegates might unbind themselves on the pretext that Trump withdrew his pledge to support the Republican nominee. Remember those chaotic Nevada caucuses that Trump won? They could be the subject of a credentials challenge. There could also be disputes over the disposition of delegates from Marco Rubio and other candidates who have dropped out of the race.3 A final possibility is “faithless delegates,” where individual delegates simply decline to vote for Trump despite being bound to do so by party rules. It’s not clear whether this is allowed under Republican rules, but it’s also not clear what the enforcement mechanism would be.

I don’t want to make too much of these “nuclear” possibilities, given that such efforts would be blatantly undemocratic and would risk a huge backlash from Republican voters. Still, even 1,237 delegates isn’t quite a safe number for Trump, especially if he’s just barely above that threshold.

Another possibility is Trump coming up somewhat short of 1,237 delegates, but close enough that he could win on the basis of uncommitted delegates who vote for him on the first ballot. In fact, Trump finishing with something like 1,200 delegates is a strong possibility. The expert panel we convened two weeks ago had Trump finishing at 1,208 delegates — with a lot of uncertainty on either side of that estimate — and he’s run just slightly behind our projected pace since then by getting shut out of delegates in Utah.

Let’s say that Trump ends with exactly 1,200 delegates after California. He’d then need 37 uncommitted delegates to win on the first ballot. That might not seem like such a tall order — there will be at least 138 uncommitted delegates, according to Daniel Nichanian’s tracking, and Trump would need only 27 percent of those. But most of those delegates4 are chosen at state meetings and conventions, the same events producing unfavorable delegate slates for Trump in Massachusetts and other states.

Alternatively, Trump could try to broker a deal with another candidate — Kasich, for example — to get to 1,237. But that isn’t so easy either; whether Kasich could instruct his delegates to vote for Trump on the first ballot would vary depending on the rules in each state, and some delegates could become unbound instead of having to vote Trump. Trump and Kasich could also try to strike a deal on the second ballot — but by that point, most of their delegates would have become free to vote as they please.

This is not an exhaustive list of complications. We’ll save the discussion about Rule 40 — and why it’s largely toothless — for another time, for instance. The basic problem for Trump is that all the rules will be written and interpreted by the delegates, delegates who mostly don’t like Trump. They have a lot of power to wield at their discretion.

That’s not to say the rest of the voting doesn’t matter — it would be much easier, both procedurally and ethically, to block Trump from getting the nomination if he comes into the convention with 1,100 delegates instead of 1,300.

Listen to the latest episode of the FiveThirtyEight elections podcast.

http://c.espnradio.com/s:5L8r1/audio/2714714/fivethirtyeightelections_2016-03-28-180814.64k.mp3?ad_params=zones%3DPreroll%2CPreroll2%2CMidroll%2CMidroll2%2CMidroll3%2CMidroll4%2CMidroll5%2CMidroll6%2CPostroll%2CPostroll2%7Cstation_id%3D4278Subscribe: iTunes |Download |RSS |Video

How Trump Hacked The Media

Donald Trump took over the news cycle on Feb. 26, as he had so many times before. In the morning, the media chatter was about Marco Rubio’s seemingly strong debate in Houston the previous evening in which he’d confronted Trump. By midday, however, there were rumors and reports that Trump would be endorsed by New Jersey Gov. Chris Christie. Once Christie, who’d been flown to Trump’s event in Texas surreptitiously, delivered his endorsement late that afternoon, Rubio was swept off the front pages and relegated to sideshow status.

I remember getting into some arguments with my colleagues that afternoon. Was Christie’s endorsement really so momentous that it required wall-to-wall coverage? You can make a case that it was: Christie was the first sitting governor and the most mainstream Republican to have endorsed Trump to that point. The case against: The endorsement was somewhat predictable based on Christie’s previous behavior toward Trump, and few of the hundreds of endorsements made during the campaign earn breaking-news chyrons.

The irony is that for such a “game-changing” event, Christie’s endorsement of Trump had a relatively short shelf life as the top news story. By noon the next day, it was just one of a mishmash of headlines about the Republican campaign; there were others about the ongoing battle between Trump and the Republican “establishment,” and even one about how a pair of women who backed Trump on social media had accused Rubio of having a “gay lifestyle.” It was mission accomplished for Trump, however, who had changed the news cycle on his whim and prevented Rubio from sustaining any momentum of his own.

For all the recent debate about what responsibility the media bears for Trump’s rise to the top of the Republican Party’s nomination race, there hasn’t been a lot of evidence presented on how the media has actually covered the campaign. So I scoured through more than nine months of headlines since Trump’s presidential bid began.

The Christie endorsement, it turns out, is emblematic of a larger pattern: Trump has been able to disrupt the news pretty much any time he wants, whether by being newsworthy, offensive, salacious or entertaining. The media has almost always played along.

But that isn’t the only problem we found.

Trump has dominated the news, but that’s not the whole story.

Sunday marked the 286th day of Trump’s campaign, which began June 16. From the start, he’s been a media phenomenon. According to The New York Times, Trump has received the equivalent of $1.9 billion in television coverage while having spent only $10 million on paid advertising. By contrast, Trump’s Republican rivals combined have received slightly less than $1.2 billion worth of television coverage, meaning that Trump has been the subject of the clear majority (62 percent) of candidate-focused TV coverage of the Republican race.

There’s a perception that Trump has dominated television coverage more than coverage in print or digital media outlets, but it’s not clear that’s true. A study we conducted in December found that 54 percent of newspaper stories about the Republican candidates were about Trump, not that far from his share of TV coverage. (For transparency’s sake: Among stories FiveThirtyEight has published where a Republican candidate’s name has appeared in the URL — which most often mirrors the headline — 43 percent have been about Trump.1)

For a further sense of how digital outlets are covering the race, we can borrow a technique I’ve used in the past, which is to record the top story as of noon each day from the news aggregator Memeorandum. The site uses an algorithm to determine which stories are leading political coverage on the Internet; the details of the calculation are somewhat opaque, but a lot of it is based on which stories are being linked to by other news organizations and what themes are commonly recurring among different news outlets. Simply put, Memeorandum is a good indicator of what stories journalists are talking about.

Through Sunday, Trump had been the lead story on Memeorandum on 104 days, or 36 percent of the time since he announced his candidacy. However, Trump is competing for coverage against not only other Republican candidates but also Democrats, along with other international and national news stories.2 Of the days when a story about the Republican campaign led Memeorandum, it was a Trump-related story 68 percent of the time.3

HOW OFTEN DID TRUMP LEAD NEWS COVERAGE OUT OF ALL DAYS …DAYS WHERE TRUMP LED COVERAGETOTAL DAYSTRUMP’S FRACTION OF COVERAGE… since the start of his campaign on June 16, including days when the top story was not campaign-related?10428636%… when a campaign-related story led the news?10420052… when a story about the Republican campaign led the news?10415268… when a story about a particular Republican candidate led the news?10413776Trump has dominated coverage of the GOP campaign, but not always all other news

Importantly, this is a measure of which topics are gaining traction rather than how many stories on a particular topic are being published. Then again, stories rank high on Memeorandum precisely because other journalists (and not just readers) are picking up on the same themes. It’s safe to say that the GOP race has been covered principally as a Trump story across all forms of media.

But how much Trump has been covered is less interesting in some ways than how he’s been covered. In fact, judging by what has been featured on Memeorandum, coverage of Trump has evolved over time, breaking down fairly neatly into three periods:

In the summer, coverage of Trump was highly poll-driven, with most of it emphasizing how popular he was despite repeatedly making inflammatory remarks.In the early fall, coverage of Trump slowed down considerably. This coincided with a period when he was relatively subdued in debates and public appearances, and also fairly stagnant in the polls. However, Trump again became a focal point after the Paris and San Bernardino terrorist attacks, which Trump used to exploit anti-Muslim and anti-immigrant sentiment.In the winter (and so far in the early spring), the coverage of Trump has been manic. He’s been the near-constant center of attention, but rarely has the media focused on any one Trump-related storyline for more than a day or two at a time.

We can consider these phases in more detail based on a series of charts designed by my colleague Reuben Fischer-Baum. They illustrate whether the lead story at Memeorandum each day was about Trump and highlight three frequently occurring categories of Trump coverage:

Polling. There have been 29 days of coverage on Trump’s polls, making it the most frequently occurring Trump-related subject.Trump versus the Republican Party. There have been 22 days of coverage on Trump’s insurrection against the Republican Party “establishment.” These stories often involve excellent reporting. Almost always, they frame the conflict as one between Trump and the Republican “party elites,” as opposed to considering divisions among Republican voters in their attitudes toward Trump.Inflammatory comments. And there have been 16 days of coverage about controversial comments made by Trump. This category relates only to comments spoken, written or tweeted by Trump, as opposed to other types of Trump-related controversies.

For a more detailed breakdown of exactly which stories were leading coverage, you can download our spreadsheet here.

The Summer Of Trump: First the media missed Trump’s popularity, then it missed his unpopularity

Here’s Reuben’s chart for the summer of Trump, which covers from Trump’s announcement of his candidacy June 164 through Sept. 22 of last year.

Trump’s entry into the campaign was initially treated dismissively — including by FiveThirtyEight — although it didn’t take that long for the tone of coverage to change. Particularly after Trump belittled Sen. John McCain’s time as a prisoner of war and many news outlets predicted his immediate demise,5 but Trump’s polls continued rising instead, Trump developed a reputation for being a “Teflon candidate” who was relatively impervious to attack.

I’ve seen a lot of commentary to the effect that the press was slow to recognize the appeal of Trump to Republican voters. Certainly there was a lot of variation among news outlets — with FiveThirtyEight being on the slower end to come around. But overall, that hypothesis doesn’t hold up well once you review the evidence. By early August, reporters were writing about how “Trump’s staying power [was] defying predictions of political doom” and how his top-line polling numbers among Republicans were evidence of his broad and surprising popularity.

Overall, in the summer, there were 13 days of coverage of Trump’s polls, more than the nine days spent on his questionable remarks.6 Furthermore, these poll-based stories were almost always7 positive for Trump. Republican voters soon took away the message that Trump was not only their most likely nominee, but would also be a strong general election candidate, frequently naming him the “most electable” of their choices.

What was wrong with the media spending so much time citing Trump’s polls — which, after all, correctly showed him with a lead? We (somewhat infamously) spent a lot of time arguing about this in the summer and fall. Part of the issue is that polls have historically not been very predictive early in primary campaigns, as candidates such as Rudy Giuliani, Howard Dean, Herman Cain, Rick Perry and Hillary Clinton (in 2008) can attest. If Trump and Clinton wind up winning their nominations this year after their early poll success, it will be the exception as much as the rule.

But the other reason is more particular to Trump, which is that the top-line poll numbers the media frequently cited don’t convey the whole polling story. Trump has consistently had the plurality of Republican support in polls, but those same polls suggest that Trump faces unusually high resistance from voters who don’t have him as their first choice. To some extent, that’s true among Republican voters. Many of them would be unhappy with a Trump nomination, more than is typical for a polling front-runner, which is part of why Trump’s path to 1,237 delegates remains tenuous. It’s more obviously and emphatically true of general election voters, with whom Trump has historically high negatives. Trump’s weakness among general election voters has been evident for a long time — it’s not a new story, even though it has just recently started to get more coverage.

Put another way, the media may have been slow to recognize Trump’s popularity — it took a month or two after he launched his campaign. But it was slower to recognize his unpopularity, and Trump is a profoundly unpopular candidate with the broader American public.

The Autumn Of Trump: A break in Trump coverage, then came Paris

There are a few other twists in our story, however. For instance, you might assume that Trump has dominated news coverage from wire to wire. But there was a long period in the early fall when the news was not very focused on Trump. In the 53-day period from Sept. 21 through Nov. 12, Trump was the lead story on Memeorandum only three times:

What was the media covering instead? There was some important non-campaign news: This period includes the narrowly averted government shutdown in late September and the nearly disastrous Republican House speakership transition in October.

Another frequent topic was the Democratic race. Even though Clinton’s polling was quite a bit better than Trump’s — she had a much larger share of the Democratic vote than Trump had of the Republican vote, a larger lead over her nearest rivals than Trump had over his, and better (although nonetheless fairly poor) general election numbers — the media usually portrayed Clinton’s polling in a negative light. There was also continuing coverage of the scandal surrounding her private email server and frequent speculation about Joe Biden entering the Democratic race. By mid-October, after a strong debate for Clinton and after Biden confirmed he wouldn’t run, the Democratic race receded from the headlines. But there’s been an interesting symbiosis between coverage of Clinton and coverage of Trump. Clinton, who has tried to run a low-key, “prevent defense” type of campaign, has probably benefited from Trump eating up so many news cycles, while Bernie Sanders has probably been hurt by it.

In late October and early November, there was also a period of coverage of the Republican race that wasn’t Trump-centric. A lot of this coverage focused on Ben Carson, who received four straight days of mostly tough coverage from Nov. 5 through Nov. 8. But the more conventional candidates didn’t receive all that much attention. From Trump’s entry into the race in June through the end of 2015, Rubio was the lead story on Memeorandum only three times, Jeb Bush also only three times and John Kasich zero times. Scott Walker was the focal point just twice — on the day he entered the race and the day after he quit it.

If the non-Trump Republicans were having trouble gaining traction, it became harder after the terrorist attacks in Paris on Nov. 13 and in San Bernardino, California, on Dec. 2. After a few days when coverage of the attacks themselves led the news, Trump made waves with his response, particularly on Dec. 7 when he called for a ban on Muslim immigration. This period coincided with a renewed rise for Trump in national polls, which had been quite stagnant throughout the fall:

Likewise, it coincided with a big increase in the amount of media coverage of Trump. From Dec. 7 through Sunday, Trump was the lead story on Memeorandum on 57 of 112 days, or just slightly more than half the time, up from about 30 percent of the time beforehand.

Although the correlation is strong, the causal relationships between Trump’s spikes in the polls and those in his media coverage are hard to sort out, in part because they probably reinforce one another. But it’s at least plausible that Paris and San Bernardino restarted the cycle of attention to Trump when it might slowly have petered out otherwise.

The Winter Of Trump: Manic coverage and panicked Republicans

Since Dec. 22, Trump has led the news more than half the time, including one streak this month where he did so for 15 days in a row. That’s partly because there’s been a lot of actual Trump-related news. Among other things: lots of debates; the mostly ineffectual efforts by Republican “party elites” to stop Trump; and — since Feb. 1 — voting in Iowa, New Hampshire and other states, which has mostly gone well for Trump.

And yet, no individual story about Trump has led news coverage for more than two consecutive days. (See here for a more detailed breakdown of topics.) Some seemingly significant stories didn’t even make it that far. When Trump canceled a rally in Chicago after clashes between supporters and protesters, it led Memeorandum for only one day. The fact that Trump has frequently condoned violence against protesters has never led a day of coverage. Christie’s endorsement of Trump led the news for only about half a day, as I mentioned. Remember when Trump got into a fight with Pope Francis? That story also led coverage for only half a day.

What stories have been missed?

You could push back against a few examples — the current lead story about the Trump campaign’s abusive treatment of reporter Michelle Fields is probably more interesting to the press than to the broader public, for instance. But most of the Trump-related stories the media has covered have a lot of intrinsic news value. It’s easy to defend breaking from Rubio’s post-debate buzz to cover the Christie endorsement, for example. The problem is that the cumulative effect of always choosing the Trump story piles up. There has been scant coverage of candidates other than Trump and Clinton — certainly including the other Republicans but also Sanders. These candidates have not received all that much opportunity to build momentum after favorable events on the campaign trail. Nor have they gotten all that much vetting or been subject to all that many investigative stories. That may be part of why Rubio’s standing fluctuated so much during February and March, for example. Voters hadn’t heard much about him: He’d led the news day only three times before February, according to Memeorandum. So a relatively minor story, such as a strong or weak debate, could weigh strongly upon their opinion of him.

Another problem is that Trump is very often dictating the terms of his coverage, both by threatening to withdraw access from outlets that treat him unfavorably and by pre-empting other stories that might be unfavorable to him. There are whole genres of Trump-related stories that remain underexplored.

According to Memeorandum, for example, at no point has an investigative story about Trump’s past business dealings led news coverage. That doesn’t mean these stories haven’t been written — there have been some good ones — but they haven’t gained traction. (And remember: Memeorandum placement is principally based on which stories are receiving inbound links from other news organizations, so we can’t just blame readers for not caring about those stories.) Also, Trump’s policy flip-flops have rarely led the news. How many Republicans know that Trump once called himself “very pro-choice,” or once promoted single-payer health care, or once called for a wealth tax? I’ve seen it asserted that Republicans don’t care very much about these things, and that may be a reasonable supposition, but the theory has never really been tested because these stories have not received much emphasis.

Trump hacked the system

Most of the media’s self-criticism of its Trump coverage has focused on whether Trump’s dominance of the news cycle reflects a craven desire for higher TV ratings or Web traffic numbers. It’s fine to debate that — although these criticisms are sometimes being evinced through crocodile tears given the record ratings and traffic Trump is bringing to news organizations of all kinds.8

But this critique avoids some thornier questions. For instance, with his ability to make news any time he wants with a tweet, news conference or conveniently placed leak, Trump has challenged news organizations’ editorial prerogative. Should the press cover a candidate differently when he makes trolling the media an explicit part of his strategy, on the theory that some coverage is almost always better than none?

That’s not the only problem. Trump also challenges the media’s notion of what it means to be “objective.” Among other things, Trump has frequently invoked misogyny and racism; he has frequently lied, and he has repeatedly encouraged violence against political protesters. As far as we’re concerned at FiveThirtyEight, these are matters of fact and not opinion and to describe them otherwise would make our reporting less objective. Other news outlets will bend over backward to avoid describing them in those terms, however.

An underappreciated problem is that Trump’s candidacy is relatively lacking in precedent, which means we’re all trying to figure this out as we go along. Traditional journalists have had trouble covering Trump, but so have empirically-minded ones like us here at FiveThirtyEight. We laid long odds against Trump getting this far, in large part because no similar presidential candidate has done so since primary and caucus voting became widespread in 1972.

Put another way, Trump has hacked the system and exposed the weaknesses in American political institutions. He’s uncovered profound flaws in the Republican Party. He’s demonstrated that third-rail issues like racism and nationalism can still be a potent political force. He’s exploited the media’s goodwill and taken advantage of the lack of trust the American public has in journalism. Trump may go away — he’s not yet assured of winning the GOP nomination, and he’ll be an underdog in November if he does — but the problems he’s exposed were years in the making, and they’ll take years to sort out.

Listen to the latest episode of the FiveThirtyEight elections podcast.

http://c.espnradio.com/s:5L8r1/audio/2714714/fivethirtyeightelections_2016-03-28-180814.64k.mp3?ad_params=zones%3DPreroll%2CPreroll2%2CMidroll%2CMidroll2%2CMidroll3%2CMidroll4%2CMidroll5%2CMidroll6%2CPostroll%2CPostroll2%7Cstation_id%3D4278Subscribe: iTunes |Download |RSS |Video