Unity Reform Commission: Superdelegates Recommendations

The following is part three in a series of posts on the recommendations passed by the DNC Unity Reform Commission during the group's final public meeting on December 8-9. While the text below matches the language presented at that meeting, there may be some subtle differences between the language here and the final report the URC will deliver to the Rules and Bylaws Committee by January 1, 2018. Those changes, if any, are unlikely to affect the intent of the substance below. Instead, the aim would be to tighten the language.

Below are the six URC recommendations for superdelegates. Any emphasis added -- bold and/or italicized language -- is FHQ's and is mainly focused in this section on any action verbs; an attempt to capture the commission's emphasis on particular goals. The vote of the commission on a given measure follows bolded in brackets. Further notes are added below each recommendation, indicating any amendments and whether/to what extent those amendments were adopted.

Part one: Primaries recommendations
Part two: Caucuses recommendations

--

Superdelegates/Unpledged Delegates Recommendations

1. The commission concurs with the determination of the 2016 quadrennial Democratic national convention that the reduction of over 400 unpledged delegate votes equaling nearly 60% of the total unpledged delegate votes in 2016 will strengthen the grassroots role in our presidential nominating process. It is important to the commission that the grassroots voice in the presidential nominating process be amplified. 

[unanimous]

2. In regards to current unpledged delegates, the commission concurs with the 2016 Democratic National Convention that Democratic members of Congress, governors, and distinguished party leaders remain automatic delegates and unpledged on all matters before the convention, and that DNC members remain automatic delegates, but that their vote be pledged [bound] only with respect to the first ballot of the presidential roll call vote. 

[unanimous]

NOTE: Weaver requested unanimous consent to change what was presented originally as "pledged" be changed to "bound" throughout the section on unpledged delegates. The request was made to clarify the difference between types of delegates. On the one hand, those delegates elected through the delegate selection process to fill a slot allocated to a particular candidate based on the results of a primary or caucus are pledged (but not bound). Weaver sought to differentiate between those delegates and the segment of superdelegates -- automatic delegates, not elected but guaranteed a delegate position -- who would be bound under the provisions of this recommendation to candidate. A candidate has the loyalty of the former group, but an enforceable binding mechanism is required to regulate it in the latter. 

There was never an objection to the request for unanimous consent. With no objection the use of "pledged" was changed to "bound" throughout the unpledged delegates portion of the recommendations (with respect to the recommended creation of the automatic but bound categories of delegates). 

3. The commission recommends that the DNC will ensure that all party officials who have a role in the execution of the actual primary or caucus process in their state must be scrupulously neutral, both in reality and perception, in their administration of electoral activities. Any person who violates this important commitment to impartiality could be subject to loss of delegate status or other privilege they may hold at the DNC. 

[unanimous] 

4. The Commission recommends the creation of three categories of automatic delegates, one of which would remain unpledged and two of which would be bound on the first ballot of the presidential roll call.

A. Category one: Democratic members of Congress, governors and distinguished party leaders who remain automatic delegates and unpledged. 

[13-2 in favor with one additional abstention; Konst and Turner in opposition]

B. Category two: state elected DNC members, DNC members elected as a state chair, vice chair, state committeeman or woman who are not part of category one would remain automatic delegates. However, on the first ballot of the presidential roll call vote, their votes would be proportionally allocated based on the outcome of the primary or caucus in the state which elected them [subject to the same thresholds that apply for the awarding of at-large, pledged delegates]. 

[unanimous with one abstention] 

NOTE: The bracketed portion in the text was added as an adjustment after public approval by the URC and revealed during reconsideration of tabled items.

C. Category three: Officers at-large and affiliated members who are not part of category one or two would remain automatic delegates. However, on the first ballot of the presidential roll call vote, their votes would be proportionally allocated based on the national outcome of the primaries and caucuses [as measured by national allocation of pledged delegates subject to the same thresholds that apply for the awarding of at-large, pledged delegates].

[unanimous with one abstention] 

NOTE: The underlined portion in the text above is an FHQ correction to what was passed as "category two or three" by the URC. The creation of the three groups here was done to comply with the convention resolution that created the Unity Reform Commission. It was the intent of the group to have a hierarchy of categories from the remaining unpledged delegates down through the two necessary types of newly bound automatic delegates.¹ Subsection C seeks to create the lowest/third category in that hierarchy, defining it by differentiating that group from the higher two categories. Category three cannot, therefore, differ from both category two and itself.

NOTE: The bracketed portion in the text was added as an adjustment after public approval by the URC and revealed during reconsideration of tabled items.

5. With respect to the actual mechanism of how these new automatic but bound categories will be allocated, the commission provides two options for the Rules and Bylaws Committee to review and adopt. Pursuant to the mandate of the 2016 Democratic National Convention, if the RBC adopts a different mechanism, this commission will review it to determine whether it constitutes a substantial adoption of the commission's recommendation. The commission anticipates that the deliberations of the RBC in this regard will be consultative in nature with the commission to ensure the joint goals of unity and reform are achieved. In drafting the two options, the commission considered but did not adopt alternate mechanisms that, while less complicated, would substantially increase the number of delegates. For all purposes, the use of the word state shall include the states, the District of Columbia, the territories, and Democrats Abroad. The Commission, therefore, provides the following mechanisms for first ballot voting for the presidential nomination to the RBC to review and adopt.

A. Pooled Vote Option

The votes of category two and category three delegates will be allocated as defined above. The state parties shall announce these allocations for category two delegates no later than when the state party certifies their pledged delegates to the DNC secretary. The DNC shall announce the allocations for category three delegates within ten days of the last nominating contest that awards pledged delegates. At the national convention, the votes from category two and three delegates shall be automatically reported by the secretary of the convention at the time of the presidential roll call vote or [and?] entered into the state tally sheet and voted in the normal order without ascribing the votes to any specific delegate.

B. Alternate Voting Option

No later than five days after the state's primary or caucus, the state chair shall poll the presidential preferences of the state's category two delegates and each category two delegate must make his or her binding preference known to the state party chair within 24 hours. If the proportion of category two delegates supporting each candidate matches the allocation to which each candidate is entitled, the state chair shall collect the written, binding presidential preference of each category two delegate and submit those preferences to the DNC. If the proportion of category two delegates for each candidate does not match the allocation to which each candidate is entitled, the state chair shall poll category two delegates currently supporting a candidate who has too many delegates and determine who is willing to change his or her binding presidential preference to another candidate who has fewer category two delegates than that to which she or he is entitled. If enough of these polled delegates agrees to change preference to provide a match to the number needed, the state party chair shall submit those changed preferences to the DNC. If the state chair determines that there is still an insufficient number of category two delegates supporting candidates proportionally, the state chair shall by lot determine which category two delegates supporting a candidate or candidates with an excess of delegates shall pass their vote at the national convention to a credentialed alternate delegate willing to support a candidate in deficit on the first round of balloting for the party's presidential nomination. Category two delegates chosen by lot in this manner shall retain all other voting rights, including the right to vote on all other matters other than the first round of balloting for the presidential nomination. Following this determination, the state party chair shall then collect and transmit to the DNC the written, binding presidential allocation of all category two delegates and alternates, if applicable, to the DNC. The DNC shall make public the identity of each category two delegate and any alternates as applicable and the candidate for which they have submitted a written, binding presidential allocation.

Those delegates who are in category three would be allocated similarly to those in category two as follows. The DNC would announce the allocation of category three delegates to presidential candidates no later than five days after the last nominating contest that allocates pledged delegates. If there are insufficient category three delegates to fill a presidential candidate's position, alternate delegates would be elevated to provide the needed presidential preferences. Once the DNC has achieved the requisite number of supporters for each candidate to match the allocation to which each candidate is entitled, the DNC shall then collect the written, binding presidential preference from each category three delegate and alternates as applicable. The DNC shall make public the identity each category three delegate and any alternates as applicable and the candidate for which they have submitted a written, binding presidential preference. 

If the state chair or the DNC determines that there is still an insufficient number of category two delegates or category three delegates, respectively, supporting candidates proportionally after attempting to allocate alternate delegates, the state chair/DNC shall still determine by lot which category two or category three delegates, respectively, supporting a candidate or candidates in excess shall pass their vote at the national convention. However, instead of passing the vote to an alternate, the vote of the delegates chosen by lot would be voted as abstain on the state tally sheet and as abstain during the roll call of the states. The secretary of the convention would then automatically announce and report a requisite number of additional votes to achieve the proportions required to reach the correct presidential candidate allocation of category two and category three delegates. 

[unanimous with one abstention]

NOTE: Both options in the subsections of section five were voted on together unlike the three subsections of section four above. 

--
FHQ Thoughts:

Hands tied: 
There never really was all that much suspense over what kind of recommendations were going to emerge from the Unity Reform Commission with respect to superdelegates. It was set in stone by the same resolution that created the group at the 2016 national convention. As FHQ wrote at the time:

While the open primaries mandate was passive, the part of the amendment devoted to superdelegates had more teeth to it. In any event, there was more clarity as to the specifics of the superdelegates mandate of the commission. While the group will broadly consider the superdelegates' role in the process, it will specifically recommend that elected officials -- Democratic members of Congress, governors and distinguished party leaders (presidents, vice presidents, etc.) -- remain as unpledged delegates. However, the second part of the recommendation shifts the remaining superdelegates (approximately two-thirds of them) out of the unpledged category and into a pledged territory (to be proportionally pledged/bound based on the results of primaries and caucuses). [Emphasis added now by FHQ]

In other words, the URC had to make this recommendation to the Rules and Bylaws Committee. They could not recommend that any additional delegate be added to or removed from the unpledged category. And while the URC was stuck with this particular recommendation, it remains to be seen whether the full DNC will stick with these recommendations as currently constructed. FHQ is not suggesting that the DNC will not adopt the measure as written, but rather that the convention resolution does not bind the DNC in the way it did the URC on the matter.


Easier said than done (part one) -- Categorization: 
The catch in all of this is that it much easier to say that a particular segment of the heretofore superdelegates will in the future have their role/power constrained than it is to actually design and implement a system that accomplishes that goal. So while the recommendation was cut and dry, the system of achieving that directive was considerably more difficult in light of the fact that there are different types of DNC members (based on how they are chosen and/or end up on the DNC). Step one in this process was categorizing in Section 4 the delegates affected by the mandated recommendation. The segment unaffected by the changes -- elected officials and distinguished party leaders (former presidents and party chairs, etc.) -- is clear enough. But different DNC members would have to be treated slightly differently in shifting from unpledged to bound. That transition requires a different mechanism for state-elected DNC members (state party chairs, for example) and at-large or affiliated members (members who hail from states and have been a part of home state delegations in the past, but who are selected/elected nationally by the DNC).


Easier said than done (part two) -- Allocation:
Categorization in place, the URC designed in Section 5 not only a mechanism for allocating those delegates (in the introductory portion of Section 5), but two options (subsections A and B) for binding the DNC member delegates in categories two and three. The allocation part of the sequence naturally follows from the categorization step. Category two (state-elected DNC members) delegates would be allocated to candidates based on results in primaries and caucuses. The method is similar to how at-large delegates are currently allocated: proportionally to candidates with greater than 15 percent of the vote in the statewide result. In other words, that would layer an additional calculation into the statewide formula to go alongside that for at-large delegates and pledged party leader and elected official (PLEO) delegates. This third, separate calculation would allocate, for example, approximately 30 percent of the category two delegates in a state to a candidate who has won 30 percent of the statewide vote.²

But category three (nationally elected DNC members) delegates are allocated similarly, albeit with a catch. Rather than being proportionally allocated based on statewide primary or caucus results, the nationally selected DNC member delegates would be proportionally allocated based on the final national results in all primaries and caucuses. The qualifying threshold would be the same 15 percent, but nationally instead of statewide.


Easier said than done (part three) -- Binding:
The first option laid out by the URC for binding, the Pooled Vote Option, borrows a concept employed by the RNC for the first time during the 2016 cycle. Essentially, the pooled vote option would do two things: 1) pool all of the category two and three delegates and 2) automatically have the secretary of the convention count their votes. This is akin to the binding mechanism the RNC instituted for the 2016 convention. And the intent in this case as in the RNC one is the same: to keep bound but conflicted delegates in line. Imagine a number of Sanders-bound but Clinton-sympathetic delegates at the Philadelphia convention, for instance, if this provision had been in place during the last cycle. Those delegates would not be unlike the Trump-bound, but Cruz/Kasich/Rubio-sympathetic delegates who were locked into Trump votes by the convention secretary during the roll call vote at the Cleveland convention.

How does the party/convention prevent mischief -- violating the binding by voting in line with the conscience -- in this type of scenario? There are two schools of thought. The first is to do what the pooled vote option lays out. It would remove the action/decision -- the ability to vote -- from the individual delegates on the first ballot nomination vote (and that vote alone). Once those delegate slots are allocated to the candidates at the time of certification of the state results (category two delegates) or once primary season is over (category three delegates), that is it. They would be added to the overall tally (and the state tallies for category two delegates) at the commencement of the first ballot roll call vote. The "who" filling that slot would be meaningless. In fact, it would be entirely withheld. The voting slots would be allocated, but remain basically unfilled; empty slots with votes automatically cast for a particular candidate on the first ballot.

Alternatively, a party/convention could employ a replacement strategy. This is what the DNC rules provide for in the case of all other pledged delegates. If, for example, a Sanders allocated slot is filled with a delegate candidate who is not loyal to -- or willing to vote for -- Sanders at the convention, then the Sanders team would have the ability to replace that delegate candidate with one who would at least be more likely to support that outcome. This is a concept FHQ raised when discussing the constraints the URC would face in dealing with a new class of automatic and bound delegates. After all, it would be a different proposition altogether to replace elected DNC members than it would be to fill allocated pledged delegate slots with loyalists as among pledged delegates now.

And while the DNC did not provide a full replacement option for the RBC to consider in lieu of the pooled vote option, it did attempt to thread the needle on a more complex alternative. The Alternate Vote Option allows DNC member delegates to retain their actual voting privileges on the first ballot nomination vote while attempting to mitigate some of the conscience and replacement issues with some constraints.

For category two delegates -- those elected on the state level -- this would work in a series of steps:

1. Poll the preferences of state-elected DNC members. If the outcome of the preference poll matches the proportional distribution resulting from the state primary or caucus, then those automatic delegates would submit a written binding agreement to the DNC. If not, then the process would move to step two.³ 
2. Volunteer alternates: If a candidate has a surplus of preference votes as compared to the primary/caucus vote-based allocation, then the state party chair would ask for volunteers from the surplus candidate's supporters to volunteer to be bound to another candidate via the same type of written binding agreement in step one above. Should fewer volunteers come forth than would be necessary to right the allocation, then move on to step three. 
3. Bound by lot alternates: If the surplus still exists, then the appropriate number of the surplus candidate's supporters would be chosen by lot to be bound by written binding agreement to another candidate to correct the allocation for the first round roll call vote. 

There is no mechanism at the secretary of the convention level to enforce the binding mechanism in a manner similar to the pooled vote option. However, the party would have the written binding agreement on which to lean, and additionally would make public the identities of the category two DNC members and the candidate to whom they are bound. That is, perhaps, a marginally softer binding process than the more severe pooled vote option.

For category three delegates, the process is slightly different. If, after the post-primary season allocation, there is a mismatch between the number of nationally elected DNC members aligned with still active candidates and the number of slots allocated to that candidate, then alternates would be elevated to correct the count.

Now, there are a couple of factors to note here. First, it is not clear from where these alternates come nor the process (if any) whereby they are selected in order to be elevated. Presumably, this task would fall to the RBC if they opt to pursue this option rather than the pooled vote or other option. Bear in mind also that these alternates would only replace the group of mismatched category three delegates on the first ballot nomination vote. The original DNC members would retain voting privileges on all other matters. And again, those mismatches represent an unknown number of additional bodies for whom the DNC would have to logistically plan. That means finding a large enough space to hold a convention and location/city that can accommodate them (and media). This is something the DNC is sensitive to since it reduced the number of base delegates by 500 between 2012 and 2016. This is likely why the URC did not recommend a similar alternate elevation process for category two delegates under this alternate vote option.

Secondly, the timing is important with respect to this process for the category three delegate allocation. It occurs at the end of primary season when typically the process has run its course. Typically. If there is one candidate remaining -- the presumptive nominee -- then this process is potentially easier to complete. Candidates who have withdrawn would have far less incentive to push for the elevation of alternates for a vote on the nomination they are not seeking.⁴ Then again, this is a mechanism for a competitive conclusion in which the nomination is unsettled heading into the convention. Even then, with fewer superdelegates, the margin would have to be within the margin of unpledged category one delegates if the choice was binary. If there were multiple candidates still vying for the nomination, however, things would be a bit messier.

And it seems like that multiple candidate convention scenario is the one that the final section of 5.B, is directed toward; the situation in which after all of the preceding maneuvering, neither the state party nor the DNC has the proper number bound category two and three delegates allocated to each candidate. Given that set of conditions, the remainder would be determined by lot but counted as abstain on the roll call tally and/or the state tally before being automatically added at the conclusion by the secretary of the convention.

One cannot say that the URC did not account for most things in these options. However, they are increasingly detailed and complex. It is easier to say delegates would be bound than to actually put a process in place to bind them.


Release the bounds: 
And what of the potential release of the newly bound delegates? The open secret about the Democratic nomination process is that all the delegates are technically free to vote for whichever candidate they choose. However, due to the manner in which the vast majority of delegates are selected -- essentially slated by the candidates (an oversimplification) -- those delegates rarely stray from the candidate to whom they are pledged. This differs from a Republican process that binds delegates to particular candidates based on the results in primaries and caucuses, but allows for the release of those delegates should the candidate to whom they are bound withdraw from the presidential nomination race.

There is no specified mechanism in the URC recommendations that accounts for that combination of possibilities. What happens to automatic delegates bound to candidates no longer actively seeking the party's nomination? The above recommendation implies that those delegates would be bound regardless of whether a candidate continues to actively seek the nomination or not. But again that is not specified. It is also unclear whether this task was punted to the RBC. However, it is something that they would likely have to consider particularly with respect to Category Two delegates. The Category Three at-large and affiliated members would be bound based on the national results rather than the state-by-state results. While Category Two delegates could conceivably be bound to candidates who cleared the 15 percent qualifying threshold, it is less likely, though not completely out of the question, that a candidate could stay above 15 percent nationally without having competed in all or most contests. But one could see that happen much more frequently on the state level, especially in the earlier states on the primary calendar. This is not an issue that will affect a lot of delegates, per se, but it is one with which the RBC will have to potentially contend when the rules baton is passed their way. That is also an area that could be a point of departure between the RBC and the URC.


Unanimity again (but with caveats): 
As with the other two sections more directly linked to the delegate selection process, the voting on the unpledged delegates section of the recommendations saw more unanimity as a rule. However, unlike the draft recommendations in the primaries and caucuses sections, there were exceptions. Yes, RBC chair, Jim Roosevelt, abstained from voting on any measure that would come before the committee he chairs. That was consistent throughout the URC voting process. But in one case there were votes of dissent. When the matter of the preservation of unpledged, automatic status for members of Congress, governors and distinguished party leaders came up for a vote, two Sanders-appointed URC members -- Nina Turner and Nomiki Konst -- voted against. The remaining Sanders-affiliated members on the commission voted for the recommended change mandated by the resolution creating the URC. This was more symbolic than anything. The recommendation still passed with overwhelming support, and despite the break in what proved to be regular unanimity, the vote signals strong support to the Rules and Bylaws Committee and the DNC.


Superdelegates versus Automatic delegates: 
The above distinctions raise another related delegate distinction that merits some discussion as well. FHQ got some pushback on social media during the final URC meeting for stating that the number of superdelegates would be reduced under the provisions of this recommendation. The major point of contention was that italicized portion. Some contended that the number would remain the same, but that in the future, some would be bound and others would remain unpledged. This is a fair point, but it raises an interesting question: What makes a superdelegates super? The above pushback suggests that the automatic delegate status is sufficient to a delegate being considered super. But FHQ would/could counter that it is both the automatic status and the delegate's unpledged status combined that make them super. And if a superdelegate loses those unfettered voting privileges on the first ballot of the presidential nomination at the convention, then they are less super or not super at all. Should the DNC adopt these recommendations as they are currently crafted, then, for the sake of clarity, FHQ will refer to category one above as superdelegates and those of categories two and three as automatic bound delegates.

Of course, the DNC nomenclature on this has been different from that generally used by the public since the party added superdelegates to the nomination equation for the 1984 cycle. The national party has always tried to keep the label "superdelegates" at arm's length; preferring to call them unpledged delegates instead. In fact, both the URC resolution and subsequently the URC have referred to them in this manner -- unpledged -- adding "or as they are also sometimes known, superdelegates" to the discussions regarding this portion of their mandate. As such, one could also simply call these new categories (automatic) unpledged and (automatic) bound and move on.


(Over)reaction:
FHQ will not belabor the point here. This has gone on a bit already. Yet, for a recommendation that was set in stone from the birth of the Unity Reform Commission at 2016 Democratic National Convention, it is worth noting that this compromise on superdelegates has been the most contentious of the lot throughout the life of the commission. No, that has not necessarily been the case on the URC itself. Again, despite the diversity of opinion on the commission, their hands were tied on the matter. The group had to make this recommendation to the Rules and Bylaws Commission.

While those various viewpoints were voiced in the context of the commission's work, most of the contentiousness was on the outside. Banners were raised for the cause of abolishing superdelegates and countered by others wanting to maintain the status quo. And that is all fine. In the context of the URC and the document that guided it, however, that sentiment -- on either side -- was misplaced. The group could have the discussion, but it was not going to affect its mandate to recommend a middle ground; a specific reduction in the number and role of superdelegates. The only lingering questions were always over implementation.

The matter -- where to come down on this superdelegates recommendation -- will ultimately be a balancing act for the full DNC. And balancing is nothing new to the DNC, or parties generally. In a recent New York Times op-ed in the wake of the final URC meeting, Julia Azari and Seth Masket spoke of a party's task of balancing the "need to win elections but also find a way to channel the different voices within their coalitions." But that second part -- the channelling of voices -- contains its own delicate balance. As Jennifer Victor described in an exchange with FHQ on superdelegates before the final round of URC meetings, "Effective democracy requires a balance of populism and elite control. Too much of one, or the other, produces suboptimal outcomes."

And that is essentially the crux of this. Superdelegates have been a part of the Democratic Party attempt at balancing populist and elite voices in the process since and during the era following the McGovern-Fraser reforms. Those who pushed through and won those reforms at and after the 1968 Democratic convention did not anticipate the proliferation of primaries that followed. Theirs was a vision of a wading into democratization, not the cannonball that states ultimately provided. To comply with the delegate selection rules changes the DNC instituted for the 1972 cycle, more states opted for the easier, state-funded primary route than the caucuses many of those behind the reforms expected. It was one of the earliest unintended consequences of the reforms.

But at its core, those changes were about the balance between the populist and elite voices in the party and the nomination process. Superdelegates were added for 1984, in part, to recalibrate the balance between the two. Not to tip in one direction or another, but rather to account for both in the process.

And although the process both internally and externally has evolved in the time since 1984, there has not been a concerted effort to revisit that balance. The issue of superdelegates has certainly been raised, but the share of superdelegates to total delegates has remained in the 15-20 percent range. And that was true even after the formula was tweaked after the 2008 cycle.

This URC recommendation is bigger than a tweak and gets at the type of change to the populist/elite balance that Victor cited. Is it too much of a lurch in the populist direction or not enough? That is the question. If the parties could go into the lab and try these things out in a dry run before the real thing they would. Instead, they are stuck with trial and error in real time during the actual primaries. That is often a road toward unintended consequences. And if one were to try to pinpoint one of those in any of the delegate selection recommendations from the URC, then the conventional wisdom of the current time would likely hold that, ahead of a competitive nomination race with no lack of prospective candidates, this superdelegates reduction is it. Is it? Will it be? Time will tell that tale.

...but the DNC will have to adopt these recommendations first.



Part one: Primaries recommendations
Part two: Caucuses recommendations


--
¹ It should be noted that there is an order to this division of automatic delegates. Should any of the delegates fit into multiple categories, then such a delegate would, for the purposes of this recommended change and its possible implementation, be counted in the highest of the categories. A delegate who is both a member of Congress and a DNC member, for example, would fall into Category One above and be treated as such rather than being slotted into one of the two bound categories.

² The at-large, PLEO and category two (automatic bound) delegates would be separately allocated in this manner -- three separate calculations -- rather than being pooled and allocated together as one statewide group of delegates. This is how at-large and PLEO delegates have been treated in national party delegate selection rules and resulting state delegate selection plans.

³ This outcome is possible, though perhaps not probable.

⁴ It is conceivable in that case that the convention votes to unbind those delegates (and perhaps category two delegates previously bound following state contests) for the roll call of states.