The first rule of the Iowa Caucus is that, unfortunately, there is an Iowa Caucus.
But look on the bright side, I’ve figured out the whole delegate allocation thing for you. It’s crazy complicated and asinine, but it wasn’t a grand conspiracy to screw Bernie.
Along the way, we’ll take a closer look at Bernie’s underperformance among people of color (POC) in Iowa (a dodgy exit poll had them voting Hilary 58-34, as kos diaried yesterday). Did Bernie’s failure to make his case to POC cost him outright victory in Iowa? Kind of. And kind of not. So impatient! First things first.
44 pledged national Democratic delegates were at stake yesterday. Right? Right.
But the Iowa clowns thought it would be fun not to elect them directly. So they invented this stupid caucus insanity. Essentially you turn up and elect precinct delegates, who are then converted to county delegates who are then converted to state delegates (cool kids call them SDEs or state-delegate equivalents) all of this through a combination of black magic and dart tossing.
I meant math. Just math. You mix the results in with other results, blend, knead, fold, and place in a pre-heated oven. And as everyone knows, when heated, piles of delegate votes shrink. So hundreds of thousands of precinct delegates become tens of thousands of county delegates which become a little under a couple thousand state delegates, all on that magical journey to the final 44.
44 what? NATIONAL DELEGATES! Remember?! That’s what we’re after here.
Why was CoinflipGate so stupid? It was over a measly precinct delegate or 0.2 of a precinct delegate, numbers that vanish into insignificance by the time we’ve melted hundreds of thousands of them together into tasty fermented mush from which to distill pure, pure state delegates. Fuhgettaboutit.
When all the smoke had cleared and the big top emptied, we had our state delegate totals: Hillary 699.57, Bernie 697.77, O’Malley 7.68, Uncommited 0.46. That’s Hillary 49.77%, Bernie 49.65%.
But Bernie still might have won. Iowa doesn’t count the popular vote, as that would be sensible.
Seriously, they don’t. Precinct delegates do not take into account how many people vote! Instead of each precinct being granted a number of delegates based on voters present, precinct delegates are divvied up ahead of time based on prior turnouts per precinct per past elections. Dumb again, Iowa!
So we’ll never really know who won the popular vote. Some stat-head might be able to figure it out by looking at the turnout numbers for every precinct and then reverse-engineering the precinct delegate allocation totals. No, I will not do that, so don’t ask.
So, actual vote aside, it’s all settled for Hillary, yeah? Ha ha, that would be too easy!
First of all, the Uncommitteds and O’Malleys could switch to Sanders if they felt like it (they won’t).
But other than that, we can just convert those state delegate totals into 44 national delegates, yeah?
Ha ha ha! No way! It’s IOWA, Jake. As this great diary by Torilahure explains, Iowa instead allocates only one set of 15 national delegates by proportion of state delegates won. Hillary 8, Bernie 7.
Then it allocates the other 29 by delegates won per Congressional district! Just for the hell of it.
District 1: 4-4 District 2: 4-4 District 3: HRC 4, The Bern 3 District 4: 3-3
As you might guess, the total pot of national delegates available in each district is based on past votes cast per District. [District 1 and 2 had equally big past votes, then #3, with #4 the smallest].
Add ‘em up and BOOM! Hillary 23, Bernie 21. But I wondered, how close was Bernie to winning? Popular vote aside, what would he have needed to tie the national delegates or win outright?
First, get past Hillary in the race for state delegates. He fell short by 1.2 delegates, or 0.12%. That’s infinitesimal. Do slightly better, and he flips the state-winner pool to 8-7 Bernie for a 22-22 split.
The Congressional District pool is trickier. According to the AP’s raw count of county delegates won in each Congressional district, we can deduce the percentage won per district by each candidate:
District 1: BS 51, HRC 49 District 2: BS 52, HRC 47
District 3: HRC 53, BS 46 District 4: HRC 50, BS 50
Had Bernie increased his score anywhere, he would quickly pick up the overall state delegate win. And national delegate tie, 22-22. Oh, caucuses. But he would have had a rough time plucking any more national delegates from a district.
Why? Because, as I learned from the tedious half-explanation of the Iowa Democratic Party’s Iowa Delegate Selection Plan and the yet more tedious Rule 13A-H of the DNC’s byzantine and convoluted caucus delegate allocation rulebook, the pot of national delegates for each district are split up by percentage of delegate vote won in each. This math just doesn’t work well for Bernie pick-ups:
District 1: Needs 5.3% more; District 2: Needs 4.3%; District 3: Needs 4%; District 4: Needs 8.5%.
Huge gaps, but any one of them would have given Bernie a net +2 gain in national delegates.
But had he won over more POC, could he have gotten beyond 21 national delegates?
It’s easy to say yes, given that any gains anywhere would have given Bernie the 8-7 statewide edge.
That would give Bernie 22-22. But to get that additional +2 flip out of any district would have taken a massive gain in POC (or any other voter demographic) support. But for the sake of argument…
Let’s take a look at how Bernie’s 58-30% deficit among POC hurt him in some of these areas.
So I did some juggling of registered Democrats in Iowa by ethnicity, recent census figures, the Guardian’s tally of the percent of county delegates won per county and the total active registered Democratic voters in each county. I was able to establish a very rough estimate of about what percent black Democrats (assuming 100% participation) were of the total Democratic voting pool in the top five counties in Iowa with the highest black population:
1) Polk, HRC 53-46, 15.3%
2) Scott, BS 51-49, 19.2%
3) Black Hawk, BS 53-47, 22%
4) Linn, BS 52-47, 10.5%
5) Johnson, BS 60-40, 10.8%
Keep in mind that these are just speculative. We don’t know what turnout rates were by demographic. But, assuming 100% turnout by all, it’s interesting to note that Sanders was able to win several counties with quite sizable black Democratic populations. As for the top-5 Latino counties:
1) Polk, HRC 53-46, 11.7%
2) Woodbury, BS 53-46, 28.2%
3) Scott, BS 51-49, 10.1%
4) Marshall, BS 54-46, 38.6%
5) Johnson, BS 60-40, 6.7%
Very strange. Even assuming POC votes going to Hillary, Sanders was able to win in many counties with quite sizable Latino voting populations as well. But the glaring exception, of course, was Polk County, whose +7 spread for Hillary mirrored her overall margin in Polk’s #3 Congressional District.
What happened in Polk, and why did it go so much worse than Woodbury or Scott for Sanders?
One ominous sign for Bernie is that Polk is home to Des Moines, a city twice as big as those in any of the other counties, by far the closest thing to an urban environment among them. If bigger-city blacks and Latinos are resisting Sanders more strongly than their small city/rural counterparts, that does not bode well for contests looming ahead.
But maybe it was also campaign strategy gone awry, too many voters and resources sent away from Des Moines to the farther reaches. And no doubt turnout, local particularities, and differing non-black/Latino populations just made other counties easier for Bernie to make hay.
It appears also that Clinton’s late “ACA repeal” fear mongering may also have been targeted and effective. 70% of POC Iowa voters said that while they felt Bernie shared their values, they trusted Clinton over him on healthcare, 35%-20%.
While it would have taken a landslide POC shift to give Bernie a national delegate win in Iowa, even a tiny bit more POC support would have given him a 22-22 delegate tie — and the ultimate bragging right of a state-delegate “win”. Looking ahead, the song remains the same. Sanders must improve with POC or he will fail to win the nomination.