An Early Look at the Byzantine Rules for Delegate Allocation in the Democratic Primary

Iowa's four congressional districts. Note Cedar Rapids in the First, Davenport and Iowa City (a college town) in the Second, Des Moines in the Third, and Ames (the other college town) in Steve King's Fourth (source).

by Thomas Neuburger

Everyone who was paying attention in 2016 remembers the Sanders-Clinton battle for delegates in the Democratic primary. In the end Clinton won that contest handily, garnering 55% of the popular vote to Sanders' 43%, and securing the votes of 2842 delegates, or 60% or the total needed to win. (The discrepancy between Clinton's popular vote percentage and her delegate percentage is due primarily to the presence of the superdelegates. The pledged delegate percentages were 54% and 46%.)

Leaving aside the volume of evidence that the DNC, via its then-chair Debbie Wasserman Schultz and various state-level party apparati, burdened the Sanders campaign unfairly and dishonestly, that contest was straightforward. Whoever won the popular vote in a given state was assigned pledged delegates roughly proportional to the winning percentage, while the superdelegates for each state went mainly to Clinton. As each contest unfolded, you could watch, count and anticipate what would happen next.

Not so this time. At least 20 candidates have started the race, and even a solid culling could leave a field of as many as ten for the start of early primaries. In addition, there's a 15% threshold that a candidate must reach in each state contest in order to receive any pledged delegates at all (Clinton failed to reach 15% in the 2016 Vermont primary, for example) AND there's a split of pledged delegates between those available based on state-at-large vote percentages and those available based on congressional district-by congressional district vote percentages.

If that sounds complicated, it is. There are also wrinkles that add to the complication (for example, Texas assigns delegates based on a senatorial districts, not congressional districts), and there are additional rules for what happens if no candidate breaks 15% in a given state or district.

The best explanation of most of this is here, "Thresholds for Democratic Party Delegate Allocation," but if you understood the paragraphs above, you have the gist, at least roughly.

To make this a little more clear, let's look at the situation in Iowa, the first-in-the-nation contest to be held on February 3, 2020.

Example: A Look at Iowa

To see the numerical breakdown between at-large pledged delegates and by-district pledged delegates, we turn to the invaluable Green Papers, in particular their Iowa page. The Democratic party has allocated Iowa a total of 49 delegates, divided into four groups as follows:

• 49 total delegate votes — 27 district / 9 at large; 5 Pledged PLEOs; 8 Unpledged PLEOs (PLEOs are party leaders and elected officials)

The total of pledged delegates is 41, the sum of the first three groups (27+ 9 + 5). The total of unpledged delegates (superdelegates) is 8. Let's focus on the 41 pledged delegates, since those are what will be available to the candidates on caucus day.

Iowa has four congressional districts, so the bulk of the pledged delegates, 27, will come from winning the district contests — about 7 delegates are available per district — with the rest (14) coming from the at-large state contest. (Since the pledged PLEOs are voted on at the state convention, I think we can assume the breakdown of this group will more or less mirror the at-large group.)

To see how that could work in practice, let's say that by the time the Iowa caucus is held, there are four viable candidates — candidates at least theoretically capable of winning 15% of the at-large or district vote in some district. Current polling in Iowa lists these candidates polling above 10%, with the rest of the field well below them:

• Biden
• Harris
• Sanders
• Warren

Let's say each of these candidates wins about 25% of the at-large vote. That divides the 14 at-large delegates roughly as follows (obviously these are made-up numbers):

• Biden: 3 delegates
• Harris: 4 delegates
• Sanders: 4 delegates
• Warren: 3 delegates

(The actual allocation math is explained here.)

Now for the district delegates: Iowa has, as noted, four congressional districts which roughly quarter the state (see map above). Three are represented in Congress by Democrats, while the Fourth, the most rural in the northwest quadrant, is represented by uber-racist Republican Steve King.

Iowa also has two college towns, one in Iowa City in the southeast district (the Second), and the other in Ames in the northwest — that is, in Steve King's district. It's major urban centers are Des Moines (in the Third), Cedar Rapids (in the First) and Davenport (in the Second).

Can Joe Biden, assuming he gets 25% of the at-large vote, also get 25% of the congressional district delegates? The First CD, with Cedar Rapids as its anchor, may go to Biden, but the two college districts, the Second and Fourth, could well go to Sanders or Warren, while Davenport in the Second CD could go to Harris.

In fact, if Biden does indeed get 25% of the statewide at-large vote, that could come from mainly rural Democrats spread out in all four districts, each of which has an urban or college center that could swing the district vote away from him. One possibility, therefore, is that Harris could win a majority of the district votes from CD-2, Sanders or Warren could win most of the district votes in CD-1 and CD-4, with CD-3 (Des Moines) being a wash.

This could leave Biden with as few as 6 of the 27 district delegates, giving him a total delegate count of 9 of the 41 pledged delegates available.

In addition, if Sanders sweeps the progressive vote in the college and urban districts (not likely, but possible), he could win as many as 20 of Iowa's 41 pledged delegates — or about 50% of them — when it's all said and done (4 at-large + 4 district delegates per district) while still winning just 25% or a little more statewide.

It could be even worse for Biden if someone like Buttigieg breaks the 15% mark in one or more of the districts without winning 15% statewide. If Buttigieg's district votes come from Biden supporters, Biden's delegate count could be even less.

The above is an extreme case of course, and the district delegate allocations may well mirror the at-large allocations in many states. Still, it's more than likely that some candidates will break above 15% in state at-large votes and still fall below 15% in many districts votes — or vice-versa, they could win only district delegates and no at-large delegates.

Either way, three things are certain. First, the overall vote in each state will not mirror the pledge delegate allocation in a great many states — a decidedly undemocratic outcome.

Second, in as close contest as this one may become, these eccentricities of allocation are likely to matter far more than they ought to.

And finally, in a close contest, the state-by-state outcomes will be unpredictable in the extreme.

Did the masters of the Democratic Party think this system would slant the victory toward the strongest anti-progressive (or faux-progressive) candidate? If so, they may have outfoxed themselves. I don't think anyone knows what this bizarre system will produce if the contest is close.