March 23, 2016
Democrats have the stage to themselves on Saturday. Caucuses in Alaska, Hawaii, and Washington weigh in on the nomination. Does it matter? We know Hillary Clinton has this virtually mathematically clinched.
Beyond her huge lead among announced super delegates, she’s ahead by more than 300 earned delegates. With the Democrats’ proportional rules, it’s the equivalent of leading a basketball game by 10 points with less than 30 seconds left.
It’s a very safe lead. However, Northern Iowa led Texas A&M by that very margin on Sunday before going on to lose in double overtime. To do the seemingly impossible, what sort of Saturday does Sanders need?
Bernie struck out in Arizona on Tuesday. It’s the type of state he has to win if he wants to get nominated. Illinois and Ohio were the type of state he had to win too. As was Missouri, which isn’t official yet, but looks like it was another Clinton victory on March 15.
If he’d won the above mentioned states, we’d find ourselves talking about a legitimate threat to the Clinton coronation. Bernie enjoys mentioning how far he’s come since he announced his candidacy with 3 percent polling support.
He cited a CBS/New York Times poll showing him trailing by 5 points nationally. While unquestionably better than where he was, it’s still less than Clinton. He’d need to win close to 60 percent of the remaining vote to catch up in earned delegates by the time of the convention.
That’s hard to pull off when you’re polling under 50 percent, even if some of the remaining states are extra favorable. Also, that was the most favorable of the recent national polls. Monmouth has him trailing by 18. The others fall in the middle.
The Real Clear Politics average has him behind by 10. Perhaps more importantly, this is slightly worse than his gap in mid-January. Bernie was about here a couple months ago, about half the states voted, he’s now trailing by a bunch of delegates, and the polls are still the same.
So why are we talking about this again?
Well, Bernie actually won more delegates last night. This doesn’t normally happen. He won Idaho and Utah while losing Arizona. You’d think the candidate who wins more states would get more delegates, but it hasn’t worked this way so far.
The weekend of March 5/6, Sanders won 3 of 4 contests. His wins in Maine, Kansas, and Nebraska were all by double digits. Two of them were almost 2 to 1. But Clinton won Louisiana, the most populous of the four, by about 50 points. That gave her an overall delegate win.
This time, his 17 or 18 point loss in Arizona was painful, especially since it was the largest of the three states, but was covered by winning about 4/5ths of the vote in Idaho and Utah. Finally, somewhere outside of Vermont, Bernie won by the same margin Hillary did in the Deep South.
If you’re trailing by double digits very late in a basketball game, you need to start hitting some three pointers. This is the equivalent.
That’s where Saturday’s contests come in. Three more opportunities for Bernie to hit a three pointer by winning by very large margins. He can only make up so much of the gap. Even if he does spectacularly, Hillary will still hold a staggering lead.
But inevitable candidates don’t usually lose 5 straight contests (counting Utah and Idaho) by enormous margins, even if they are caucuses in states often a bit off the radar. A 60/40 margin won’t cut it. A 2 to 1 margin won’t cut it. Especially in Hawaii and Alaska, he needs 75/25, 80/20.
Part of it is the math. He needs this to avoid numerical ineligibility. The larger part is narrative. Wisconsin is up on April 5. It’s a larger primary state, but potentially very favorable to Bernie. Wyoming caucuses on April 9.
If he can win 75 to 80 percent in Idaho, Utah, Hawaii, Alaska, and Wyoming, and 60 to 70 percent in Washington and Wisconsin, it’s a solid run. It’s not enough by itself. Bernie would enter New York on April 19 down by 8 points with 20 seconds left.
That’s almost a month from now. In that time, Hillary would need to slip up, the contours of the race would need to change, Bernie would need to start convincing a higher percentage of Latino voters than he has so far. Still a tough project, but better than being totally eliminated.
First he needs three landslides on Saturday. We have very limited data. Nothing in Washington. Nothing in Hawaii. One dated poll in Alaska. Hillary led 44/41 back in January.
Normally, you would figure this means Bernie couldn’t possibly win by a wide margin. Except he was only ahead 52/44 in a recent Utah poll, and 47/45 a month ago in Idaho. These are states where he flirted with 80% when the actual vote happened.
Caucus polls have consistently badly underestimated Bernie’s support. With the results from last night, we now have 4 of Ross Perot’s best 5 states in 1992 to look at. The closest of the group was Maine, where Sanders won by 30.
Alaska is the missing one. We’ve seen the same effect with Ted Cruz, and he completed his sweep of the 5 best Perot states by winning almost 70% in Utah last night against two opponents. All of this points to Bernie getting the result he needs there.
What about Hawaii? This one is less definite. It was not a good Perot state. The Texas Independent received just under 19% of the nationwide vote in 1992. In Hawaii, it was 14%.
How predictive is the Perot factor? Here are the states to have voted in 2016, in order of Perot support:
Maine 30.44% (Big Sanders win)
Utah 27.34% (Huge Sanders win)
Idaho 27.05% (Huge Sanders win)
Kansas 26.99% (Big Sanders win)
Nevada 26.19% (Narrow Clinton win)
Minnesota 23.96% (Strong Sanders win)
Arizona 23.79% (Strong Clinton win)
Nebraska 23.63% (Strong Sanders win)
Colorado 23.32% (Strong Sanders win)
Oklahoma 23.01% (Medium Sanders win)
Massachusetts 22.80% (Narrow Clinton win)
Vermont 22.78% (Huge Sanders win)
New Hampshire 22.59% (Strong Sanders win)
Texas 22.01% (Big Clinton win)
Missouri 21.69% (Very Narrow Clinton win)
Ohio 20.98% (Strong Clinton win)
Florida 19.82% (Big Clinton win)
Michigan 19.30% (Narrow Sanders win)
Iowa 18.71% (Tie)
Illinois 16.64% (Narrow Clinton win)
North Carolina 13.70% (Strong Clinton win)
Virginia 13.63% (Big Clinton win)
Georgia 13.34% (Huge Clinton win)
Louisiana 11.81% (Huge Clinton win)
South Carolina 11.55% (Huge Clinton win)
Alabama 10.85% (Huge Clinton win)
Arkansas 10.43% (Big Clinton win)
Tennessee 10.09% (Big Clinton win)
Mississippi 8.72% (Huge Clinton win)
In case you were wondering, a narrow win is mid-single digits or less. A strong win is safely into double digits. A big win is approaching a 2 to 1 ratio. A huge win is on the order of 50 points.
There’s a definite correlation. Other factors apply. Caucuses are better for Sanders. If it’s a primary, open works better than closed. He outperformed in his home state of Vermont, and underperformed in Perot’s home state of Texas.
Sanders hasn’t won a single state which gave Perot less support than the national average. However, he’s won almost every caucus. Among the caucus states, only Iowa was not well above Perot’s national average.
This presents a bit of a problem. Is he winning because they’re caucuses or because they’re Perot states? Or is there something about a state that makes it both Perot-friendly and more likely to hold a caucus?
Moving forward, the remaining caucus states, places like Wyoming and the Dakotas, were pro-Perot. Hawaii is an extreme outlier as an anti-Perot caucus state. Bernie is very likely to win, but the above numbers indicate his odds of doing so by the margin he would prefer are low. It will help settle our chicken and egg argument.
He has the endorsement of popular local Congresswoman Tulsi Gabbard. In Minnesota, Bernie was supported by her peer Keith Ellison and did well. In Arizona, Raul Grijalva was less useful. The first was an open caucus, the second a closed primary.
The states were almost identical on the Perot Index, so the open caucus/closed primary difference would account for most of the opposite result.
Washington supported Perot at the same level as Arizona and Minnesota. It’s an open caucus, the best possible arrangement for Bernie. He’s very likely to win by at least a 60/40 margin, but can he get further? Hillary had her election night rally in Seattle in an attempt to close the gap.
I’ve avoided mentioning ethnicity. We know Bernie does better with white voters than African American or Latino. His two defeats in the higher reaches of the Perot Index are heavily Latino.
Hawaii is heavily Asian or Pacific Islander. Washington has plenty of Asian voters too. We don’t have much to go on from previous contests.
Does Sanders struggle with all voters of color over a certain age or just African Americans and Latinos? Saturday will help us find out.
Bernie will win Alaska by a ton, but needs three huge landslides for momentum, narrative, and delegate math. It appears the other two states are a significantly tougher task. We haven’t seen much change in the position of the two candidates since voting began.
Most previous outcomes make sense if you account for the few main variables. If Sanders wins all three by 40 to 50 points, it would show actual progress. Otherwise, it’s always nice to win and pick up delegates, but we’ll end Saturday where we started it.