Month: March 2016

Trump needs New York, California, and Unbound Delegates

Our model takes into account the state-specific delegate selection rules and the state-specific poll averages (if they’re available). In roughly half of the plausible scenarios we run, Donald Trump comes up around 100 delegates short of the 1,237 he needs to win on the first ballot. In real life, this probably requires more strategy than #NeverTrump actors have heretofore realized, but makes Cook Political/FiveThirtyEight’s “two-front war” look like what the electorate is already planning.

If Trump fails to exceed his poll numbers and Kasich wins districts in which a Democratic representative won by more than 10 points in 2014, Trump gets 1,147 pledged delegates.

The chart above is a reasonably likely outcome: Trump fails to exceed his poll averages and John Kasich wins congressional districts in which the House member won by over 10 percentage points in 2014.

TrumpCeilingMarch29

Above is the math behind that projection. A couple of notes: Colorado is just a guess for delegate outcomes, the state will be a black box. The model has Trump winning 100% of the Congressional districts, in New York’s district winner-take-all contests based on the strength of his state-wide polling (he’s a good 35 points ahead of Kasich), but it’s hard to imagine Ted Cruz and Kasich failing to pick up. That’s probably balanced by Katich’s stronger showing in California, where the model shows him picking up 3 delegates in 30 of the 53 House districts.

Closer examination shows a big gap in Pennsylvania’s delegates. That’s because voters select unbound delegates on the district basis. West Virginia follows a somewhat similar model, where delegates may or may not declare their intention on the ballot. There are already 75 unbound delegates heading to Cleveland, and another 174 bound to candidates who have “suspended” their campaigns. That’s a big wildcard: enough to push Trump over the edge if a majority break for him.

This outcome actually requires Kasich to stick around and remain viable through June. Kasich can’t win the 1,237 needed on the first ballot; there just aren’t enough points left on the board, so a vote for Kasich is just a “none of the above” vote. Sure, he beat Trump in Utah, but he lost to Rubio – who had dropped out a week earlier – in Arizona due to the prevalence of early voting. It isn’t too big of a stretch to see the “mathematically unviable” argument swaying enough voters in Pennsylvania, New York, and California to make his spoiler status not that big of a deal.

It’s also not a sure thing that this is the direction late-deciders break. The models above show Trump picking up 0% of undecided voters. If Cruz picks up 60% and Trump picks up just 20%, the race looks a lot different:

CruzMomentumMar29

In that projection, Trump falls just 3 delegates short, and it’s easy to imagine him picking those up among the unbound delegates.

If, based on that slightly weaker projection for Kasich, he loses more of the Democrat-heavy congressional districts in California, with Trump and Cruz picking up a proportional share, Trump wins outright:

Cruz Momentum with no Kasich CD bonus

None of this takes into account the political realities that govern the later elections and a brokered convention itself. As Trump has picked up some endorsements (and Cruz only some half-hearted ones), GOP leaders could fear the “riots in the streets” outcome, or fear that by suppressing Trump the party could permanently lose the voters he’s won. It also assumes that almost no delegates jump on the Trump bandwagon in subsequent ballots. Nevertheless, it shows there’s still a reasonable set of outcomes in which Trump fails to secure 1,237 delegates before Cleveland, if the GOP plays its cards right.

Big Sanders Wins Coming, But Trouble Looms In New York

Bernie Sanders is likely to rack up a string of wins over the next two weeks: it’s a favorable calendar combining both caucuses and voting demographics that have favored him throughout the cycle. It’s entirely plausible that he wins the next five contests — most of them in blowouts.

In so doing, he may chip in to Hillary Clinton’s lead by 70-80 delegates, leaving her around 250 delegates ahead, and helping forestall for a bit the growing narrative that he doesn’t have a chance. We should note here that there isn’t much data in the upcoming contests, but our upgraded projection model allows us to posit what a Sanders blowout might look like.

A favorable calendar helps Sanders chip away, but a strong Clinton lead in New York erases his progress.

You can see above that there’s no polling data for Alaska, Hawaii, or Washington state, so we’ve plugged in big wins for Sanders (with Wyoming following suit). That would track with similar performance in small caucus states with similar demographics.

The potential trouble for Sanders, though, is that the latest polls in New York show him trailing badly. The 67-24 split would be among her bigger margins. Even if she were to drop to a more-realistic 55-45 win, the two Democrats would break even on delegates over the six contests, and Sanders will be left to fill a 310-delegate gap with significantly fewer chances.

The silver lining for Sanders is that our projection shows Clinton falling short of the pledged delegates she’s need to win the nomination outright, meaning she would need around 15% of the superdelegates to lean her way (she has nearly two-thirds in her camp now). Barring a big turnaround in New York, though, Sanders is unlikely to make a dent in Clinton’s lead.

Model updates

We’ve completely overhauled the projections spreadsheet on the Republican side to narrow in on the variables that are likely to matter in the 18 contests before the convention. The updates also allow for more-granular control to adjust the remaining states on both sides, which will be especially useful due to the lack of state-level polling data for the upcoming states. That granular control is not user-editable (yet), but we’ll be using the scenarios to help make predictions in the coming weeks.

We’ll have a series of posts in the coming days, but the updated model suggests that Donald Trump’s path to 1,237 delegates requires strong shows in California, New York, and Indiana. Our projection still shows him falling short, needing to pick up a significant portion of the unbound delegates (and Marco Rubio’s soon-to-be unbound delegates) to avoid a second ballot.

Quick take: Another Tuesday

We’re updating the underlying model to tune it a bit and allow a little bit greater variable control, but it won’t be ready until after the March 22 primaries. These may end up being relatively straightforward.

Republicans

Surprisingly, this side works out pretty easily. Arizona is winner-take-all. Trump is leading Cruz by over 10 points, and should all 58 delegates. Utah offers a bit of excitement. Cruz is polling just over 50%. Our model has him taking nearly 55%. If he gets above 50, he gets all 40 delegates. If not, he has to split them proportionally with Kasich (who would get 10-12) and Trump (who’d get 5).

Democrats

The next week should be good for Bernie Sanders.

Sanders should take at least 50% of the next week's delegates...but that does little to close his gap. — Sanders should take at least 50% of the next week’s delegates…but that does little to close his gap.

That’s a bunch of caucuses in states that should favor him. The one where he might come out behind is Arizona. There’s simply no state-level data in most of the states though. One big pickup for him could be Washington, where a 10- or 20-point gap could help him chip into Clinton’s delegate lead. In the best-case scenario, though, it’s hard to see him making up more than 30 delegates all in, which is less than 10% of what he needs to get even. April may help, but new polls in New York show Clinton building a big lead there.

Does Bernie have a chance? Probably not.

We were surprised to see a lot (a lot) of social media stories yesterday suggesting that Bernie Sanders remains strongly positioned to win the nomination. Our favorite was a post saying Sanders “won nearly half of the delegates” on Tuesday night, and, that with him only needing 1570-odd delegates to hit the 2,383 magic number and there being 2,404 on the board, the momentum was there. There were also a bunch of comparisons to Barack Obama’s path in 2008.

Poppycock.

As always, Nate Silver has a strong and succinct analysis of where things stand. It’s not impossible for Sanders to pull out a win, but for him to do so there would have to be a major shock to the system. A much more likely outcome would be that he could prevent Hillary Clinton from hitting 2,383 before the convention (i.e. without superdelegates), possibly undermining her legitimacy as the nominee – but that still has her beating Sanders by around 600 delegates.

Now let’s poke holes in this narrative. Here’s what our projection is based on:

The projection has Clinton adding another 300 delegates to her already more than 300 delegate lead. It includes some (sometimes very old) state polling in Utah, Wisconsin, New York, Pennsylvania, and California. This is the “Bernie takes majority of undecideds, but not all” model that has played out over the past few weeks. This still probably overplays Hillary’s strength a tad (there’s a string of caucus states over the next four weeks, for example), which could help Sanders pick up maybe 10% better than he stands.

Another weakness in the model is that there is precious little polling, and the demographics of many of these states is at variance to those that have already voted – but it’s not totally silly to see Hillary coming out around 50% in most states.

Here’s what it looks like if Sanders picks up all undecideds:

Hillary's ceiling is her poll average, Bernie picks up the rest. — Hillary’s ceiling is her poll average, Bernie picks up the rest.

That green line at the tippy-top is 2,383. She misses, but still leads.

In order for Sanders to catch Clinton, he would need to invert the poll numbers in the chart above. That means on average, he’d need to win all remaining states 55-45. In that scenario, both would fall short of the nomination without superdelegates. It’s not mathematically impossible, but it doesn’t track with where things are historically.

Will Clinton get 400 delegates tonight?

Before voting opened, we said that 400 was Hillary Clinton’s “over perform” number, expecting her to pull in closer to 368 delegates. We’ll update the delegate counts in the morning, but it looks like she’ll hit that 400 target. Combined with the narrative of winning 5 of 5 states (if only just), she would seem to be back on track to get the 2,383 delegates see needs without dipping into her superdelegate reserve. She should have a 300-330 delegate lead when she wakes up tomorrow. For context, Sanders would need to win every delegate in the next seven states just to catch up. He needs blowouts in places like Wisconsin and Washington to even start chipping into the lead.

Does Trump get 1,237 delegates?

It’s close, but it looks like Donald Trump will win Missouri, getting the 12 bonus delegates that go with it. We have some model updating to do in the coming days, including adding elements that help lean low-count delegate states to the right candidate.

In the meantime, now that some of our projections are now fact, we ran a few versions of the model to see what the future might hold. We think a three-way race probably hurts Trump the most, with blue states (like Pennsylvania and Maryland) giving Kasich a good chance and “values” states (like Utah) helping Cruz. We don’t know what to make of Indiana, home to many religious conservatives (but also Kokomo).

Here is our (rough) look ahead. It doesn’t factor in momentum, or states that don’t have local polls but may deviate from national based on past voting patterns. It does suggest, however, that Trump winning 1,267 delegates is far from a lock.

Depending on how Rubio voters break, it's far from a lock that Trump gets 1,267 delegates. — Depending on how Rubio voters break, it’s far from a lock that Trump gets 1,267 delegates.

Early returns: Trump’s ceiling and Sanders’ undecideds

It’s early still, but the returns look to show that Trump will come out pretty close to our projections, though may lose a few delegates in Illinois and a few more in Missouri. It’s limited data, but the returns suggest Trump’s ceiling correlates closely to his poll numbers. With that and Rubio’s drop out, our model shows Trump missing a win by around 150 delegates – particularly if Rubio (and perhaps Bush) throw some weight behind Kasich. Big asterisks: Kasich still might not end up on Pennsylvania ballot.

On the Democratic side, Clinton appears to be exceeding our projections slightly, particularly in Ohio and Illinois (and, of course, Missouri, where we had her losing by nearly 10). This suggests her polls need to add somewhere near a one-third take of the undecideds (vice the worst-case 0% we projected). This still has Clinton short exactly 100 delegates from an outright, no-superdelegate win, but she finds herself in a much stronger position.

A Note on Illinois

We made an error in our model. Because of the early assumption we made that congressional districts would break, on average, as the statewide total did, we may miss (and miss big) on Illinois delegate counts. As a shortcut, our model made it a GOP winner-take-all contest. In fact, it’s winner-take-all for the 12 statewide delegates, and winner-take-all in each Congressional district. We still think Trump will get the majority there, but Cruz and Kasich may chip into his count.

This will be corrected when results come in, and shouldn’t have too big of an impact on the overall projection. Apologies for the error.

Beware the Ides of March: Candidate Targets

There will be a new narrative in each race after tonight’s primaries. If Trump takes all of the winner-take-all states, it’s more likely than not John Kasich and Marco Rubio will drop out. Establishment types will likely view that a Hobson’s choice, and rumors of a third-party candidate may grow. Below is the Donald Trump worst-case scenario, in which his poll numbers represent ceiling. This is not our projection, but rather a baseline to measure Trump’s performance against tonight (and we left the “force Cruz win in IL” flag on):

Trump still wins the day, even if his poll numbers are a ceiling, but Kasich wins Ohio and Cruz pulls out a surprise in Illinois.

The biggest surprises above are Ted Cruz’ strength in Illinois and Missouri. Western Illinois’s not all that different from Eastern Iowa (nor, for that matter, is Northern Missouri), so Cruz doing well in those regions won’t be too big of a surprise. Chicagoland has a lot of hard-over ex-union voters, and south of Springfield, IL it’s basically Kentucky (ditto Jeff City). Our “status quo” projection, on which our current overall projection (Trump wins on first ballot, but by a nose) is based looks like this:

GOP Rule of Thumb: Measure against Trump. If he gets fewer than 150 delegates (winning only Florida and North Carolina^[1]As of this writing, he already won the Northern Mariana Islands’ nine delegates. There’s zero data there, but we wonder how much the Islands’ strange history and current legal framework for immigration affected the thoughts of the (likely) couple of dozen voters.), he had a rough night, and the likelihood of a contested convention grows; between 150 and 300, the status quo is preserved; over 300 he had an outstanding night, and he likely wins outright sometime in May.

On the Democratic side, it’s increasingly likely Bernie Sanders will upset Hillary Clinton in some key states, but due to the proportional manner in which delegates are allocated, her delegate lead will still grow. Below is the math if her poll numbers represent her ceiling:

Close States + Hillary ceiling = tough night for the frontrunner?

It’s much more likely that Illinois and Missouri will look more like Ohio, with the winner taking a smaller delegate lead than the above chart suggests. Still, it shouldn’t be a huge surprise if Bernie takes 3 of 5 states. If the numbers above hold, Sanders is on track to finish just 250 delegates shy of Clinton, but she would still com out on top with superdelegates.

Democrat rule of thumb: Measure against Clinton delegates. States won means very little to the math ^[2]Though admittedly a lot for narrative, which affects future states.. If Clinton wins fewer than 340 delegates, she’s in real trouble, and Bernie may well overtake her in delegate counts by June; if she gets between 340-365 delegates, the Michigan momentum swing is real; if she nets between 365-385, the status quo is maintained; anything above 385 is a pretty good night for her. If Clinton gets more than 400 delegates, the tightening over the past week is mostly erased and she’s back on track to win without resorting to superdelegates.

Footnotes [ + ]

1.	↑	As of this writing, he already won the Northern Mariana Islands’ nine delegates. There’s zero data there, but we wonder how much the Islands’ strange history and current legal framework for immigration affected the thoughts of the (likely) couple of dozen voters.
2.	↑	Though admittedly a lot for narrative, which affects future states.