January 10, 2016
So this is hard. Very hard. But not impossible. In fact, I’m willing to assign 12% odds.
Step 1: Win Iowa. A 50/50 proposition. Hillary is ahead right now (pending the first post New Year polls), but not by an insurmountable margin. Ann Selzer, the most widely respected Iowa pollster, most recently had Bernie within 9.
Bernie should peak at the right time, and many of the younger voters he needs to turn out managed to caucus once before in 2008. If he does surge, that means a few Hillary leaners flip.
Combined with O’Malley voters needing a home when he falls short of the 15% rule at individual caucus sites, that gets him there. It feels like this is sort of binary. Either all of the above fall into place for Bernie, or none, hence the coin flip. His momentum will impact where the O’Malley people go.
Step 2: Win New Hampshire. Real Clear Politics has Bernie ahead by a few points in their average, but Hillary is actually leading in 3 of the 5 most recent surveys.
Bernie’s two bright spots are an online NBC/Survey Monkey poll where he is up by 14 and the new Fox poll.
Generally speaking, phone polls are considered more accurate, but Survey Monkey apparently did well on the British elections.
Where Hillary leads, they’re very close. With those and the RCP average within the margin of error, they’re effectively tied now.
An Iowa win gives an average bump of 7 points to the winner. Assuming Bernie triumphs there, he’s 80/20 to win New Hampshire, moving his combined odds to 40% (.5 x .8).
You can also see how an Iowa loss would likely doom him.
Step 3: Win Nevada. If the Nevada caucus was today, you would favor Hillary. Bernie has made notable progress over the last few weeks, has positive underlying poll numbers and plenty of ground support.
She’s still ahead. But that’s Mostly Inevitable Hillary, not wounded front runner who just lost Iowa and New Hampshire. Victories in the first two contests would do a few things:
First, it would continue to motivate Bernie volunteers. You’d have to assume an endless supply of SoCal Berners helping out. Normal caucus turnout is very low. An extra 10,000 people would noticeably change the outcome.
Second, he becomes extra viable to those who like him better than Hillary, but don’t want to support a lost cause. Same deal as Iowa, but with two positive results already banked.
Third, more union support, or at least lack thereof for Hillary. The Culinary Workers Union, most important in Nevada, has stayed away from endorsing so far. Having supported Obama last time, they could jump in for Bernie. At a minimum it would prevent them from endorsing Hillary.
In terms of ground operation in populous Clark County (home of Las Vegas), Sanders is arguably ahead of Obama’s pace from 2008. His slight delegate edge came from surprising Hillary in rural areas. She’s ready this time.
Bernie has a real chance to win the popular vote, which Obama failed to. He just shouldn’t expect to have the delegate count bail him out. I’m counting a victory as finishing ahead in both, so that there’s no question who won.
With Hillary still well ahead (for now) among non-white voters, even with the momentum, I don’t want to assume Bernie’s chances are better than 60/40 to run his streak to 3 states.
He was at 40% odds after New Hampshire, multiplying by .6 and he’s now at 24% chance to sweep the first three. Had Obama done this, he would have virtually clinched the election.
This is different. Hillary has all the endorsements, won’t be broke this time, and South Carolina isn’t one of the three victories. That probably is a legit firewall for her. In the best poll for Bernie, he’s trailing by 36 points. In 2008 it was the beginning of her end.
Bernie would close the gap some with three consecutive wins, but assume Hillary wins this one, pretty much no matter what. She isn’t going to give up, and the schedule immediately following South Carolina favors her. Super Tuesday is angled toward the South, and she has a clear advantage there.
With delegates assigned proportionately, the race would drag on for a very long time, much as it did in 2008. Hillary has huge institutional advantages, but Bernie should have a very motivated base and almost limitless ability to raise money from small donors.
So another 50/50 proposition. Multiply against the 24% chance Bernie wins the first three and you have my 12% odds of winning the nomination. This is a little more optimistic than average. Consensus right now is closer to 5-10% but it’s in range.
Let’s work our way through the calendar and see which states Bernie would need to win, and which states Hillary would look to build a large enough margin to keep him from his share of delegates.
That’s still his biggest danger, that Hillary wins some Southern states by more than he can win any of his. That would give her more total votes and more delegates, even if he won as many or more states.
We can safely assume the following state characteristics help Bernie:
Independents allowed to participate in primary/caucus
This one is simple. Hillary is doing better among registered Democrats, Bernie with Independents. If a state has a fully closed primary, one that only allows Democrats in, this is a big advantage for her.
New Hampshire is an opportunity for Bernie as much because of the ability of Independents to opt in as due to the favorable geography. Some primaries are completely open, allowing voters registered with the other party to cross over.
That may help Bernie too, especially in the South, where he’s otherwise going to have a real problem staying close enough to grab a decent amount of delegates.
There are plenty of Southern conservatives who are still registered Democrats. Especially among older voters, registration is a remnant of when Democrats controlled the region. These voters now vote GOP in general elections with great regularity.
These voters, definitely not liberal progressives, are not likely to favor Sanders in huge numbers. Having them bolt for the other primary wouldn’t hurt.
Past history of supporting insurgent candidates
Some states have a more rebellious streak than others. If the state voted for McGovern in 1972, that’s a helpful indicator. Hart in 1984, not bad. Even a strong result for Ross Perot in the 1992 general election isn’t a bad signal.
Obama established himself as a serious mainstream candidate quickly, and did better with African-Americans than Sanders can hope to, so while it makes sense to look at his results, they aren’t necessarily predictive.
Similarly, Hillary is positioned a little differently this time too, so just because she won last time doesn’t mean we should assume she has a big advantage this time.
While I would expect Iowa and New Hampshire to wind up competitive like in 2008, last time Clinton did better in Nevada than South Carolina. I would expect the reverse this time, regardless of who wins them.
Voters skew younger than average
As we know, the younger the voter, the more likely they Feel the Bern. Voters under 40 heavily favor him. Voters over 65 heavily favor Clinton. In between, Clinton has a definite edge right now, but there are more undecided.
Higher percentage of white voters, at least in the first several weeks of voting
There isn’t any long-term reason why non-white voters should object to Bernie. As we saw in part one, they don’t, holding mostly favorable views with some still undecided.
But Hillary is unquestionably ahead for now. Given enough time, the gap in voting support may shrink to match the favorability difference, but we shouldn’t assume this is going to happen quickly enough for the March 1 primaries.
Caucuses require more commitment than primaries. Any Sanders win is based on early successes firing up the troops and having more energy than the Clinton side. The best place to leverage this is in a caucus, something Obama used to considerable benefit in 2008.
As mentioned above, Hillary won’t forget the delegate count this time, so he can’t rack up the same margin Obama did. Bernie is still better off in a caucus than a primary though, all other things being equal.
Candidates go on winning and losing streaks. It happened in 2008, when Obama swept several states in a row after Super Tuesday, establishing himself as the clear leader. If a state that could go either way is voting soon after a stretch that looks good for Bernie, that helps too.
All of this is a matter of degree. Very few states are 100% favorable or unfavorable on these metrics.
While Sean Trende at RCP has build a fairly detailed delegate calculator for the GOP side, allowing you to plug in your projected vote percentage in each state and have the tool do the work for you, the seeming one-sidedness of the Dem race has prevented similar work being done for this.
This means we get to cheat.
I’m going to total up the amount of pledged delegates in the states Bernie would need to win. Then we compare to the same for the states Hillary holds on to even if everything works out well for him.
Given her edge in public endorsements from super delegates, and chance of winning a place like Texas by a pretty big margin, he needs a noticeable advantage in this count to win.
Here’s an example from the first four states:
Iowa (46) Bernie
New Hampshire (24) Bernie
Nevada (31) Bernie
South Carolina (51) Hillary
Our system has Bernie ahead 101 to 51, with the knowledge that it’s closer than that looks. In reality, they would find themselves just about tied in delegates.
Hillary gets Alabama (52), Arkansas (32), Oklahoma (38), Georgia (98).
I can’t figure out how Bernie wins any of these. They are all open primaries, so it’s possible some centrist to conservative Democrats could take the opportunity to vote on the GOP side. I’d expect it. Not enough.
Oklahoma is arguably the most conservative state in the country. Unless you think it’s Alabama. Georgia saved Mondale in 1984, and has a high African-American population.
None are real heavy on college towns, none are caucuses, and Hillary would have won the most recent contest in South Carolina.
Bernie gets Colorado (62), Minnesota (78), Vermont (15)
If he can’t win these, we have nothing to discuss here. I didn’t list this above, but another factor is weed legalization. If it’s legal, it’s good for Bernie. Colorado is a decently young, insurgent supporting, caucus state where everyone is high, regardless of altitude.
Can’t you see all classes closed in Boulder so everyone has time to caucus for Bernie? Ok, maybe not, but you get the idea.
Minnesota is educated, white, and voted for Al Franken. And Jesse Ventura. It’s an open caucus. Short of having the proceedings sponsored by High Times, and open only to voters under 29, you couldn’t come up with a more Bernie-favorable environment.
Advantage Hillary: Texas (208), Massachusetts (95), Democrats Abroad (13), American Samoa (4)
Hillary narrowly defeated Obama in Texas. The denizens of Austin were just as likely to prefer Barack as Bernie, so he shouldn’t expect a huge improvement there.
I don’t want to expect Sanders to run ahead of Obama with non-white voters in the rest of the state. More likely Bernie would worry about Hillary grabbing a disproportionate amount of delegates.
The only reason this isn’t in the definite column is an open primary with Trump and Texas Senator Cruz being a draw for moderate to conservative Democrats. Odds are really thin though.
I don’t have the energy to explain my decision on Democrats Abroad and American Samoa. You don’t have the energy to read it.
You might think Bernie should have a shot in Massachusetts. It does border Vermont after all. If Elizabeth Warren endorses him before March 1, maybe, maybe it’s more of a toss-up. Hillary easily beat Obama in 2008 after he was endorsed by the Kennedy family.
I repeat. Hillary easily beat Obama in 2008 after he was endorsed by the Kennedy family.
Mass Independents are just as likely to turn out for Trump as Bernie, if not more so. There are about seven registered Republicans in the state, yet GOP candidates are sometimes elected statewide. It’s with those Independent votes. There’s a scenario where Bernie wins here, but I’m not counting on it.
Bernie Needs: Virginia (95), Tennessee (68)
As you’ll see below, even with these two, he’s ending the day trailing overall. Virginia is in play. It’s educated, and Obama won by a 2:1 margin in 2008. Even if Clinton does better with African-American voters, it’s not enough to close the gap.
Hillary had the benefit of some rural, Jim Webb-type voters in ’08. Bernie will do better with them than Obama did, so I’m comfortable giving this state to Sanders. Others will vote for Trump (open primary).
Remember, this is after he already won Nevada, so he looks more viable in the scenario than today. The one you can really argue with is Tennessee.
Hillary won this last time and that’s with an adequate amount of African-American voters who supported Obama last time but may lean Hillary this time.
But. Super Tuesday was bigger in 2008, including places like California. As organized as Obama was, he couldn’t focus equally everywhere. His priority was caucus states where he could pick up a big delegate margin.
Upsetting Hillary in the Golden State would have dealt her a crushing blow. He simply didn’t put the resources into the Volunteer State that went elsewhere, which made sense.
This is not just another version of Alabama. Tennessee is Mid-South, not Deep South and has a liberal streak to it sometimes. Nashville was a big recruiting ground for the civil rights movement. Arguably the first modern insurgent candidate was Senator Estes Kefauver of Tennessee, who knocked out President Truman in New Hampshire in 1952.
Yeah, that’s a really long time ago, but Sanders needs this one to make the math work, something that has also occurred to his staff. They will wind up ignoring about half the states on the March 1 board, allowing them to concentrate resources here.
The open primary doesn’t hurt either. I think he pulls it off.
Scoreboard after March 1: Bernie (471), Hillary (643)
This is why Bernie has to win the first three. Even with all of that in his favor, momentum, winning all the states I can reasonably give him on March 1, he’s still trailing here.
If you’re being neutral/objective and disagree with how I’ve assigned the states, if anything, you’re thinking I’m being to optimistic for Bernie and would give Hillary Tennessee or something.
Can he catch up? Yes. See how in Part Three.