On the NPC’s Decision to Remove the All Member Poll from the 2028 Presidential Endorsement Process: What Happened, Why, and What’s Next
By Olivia M, Joe W, Kareem E, Frances G, and Cara T
We are writing as 5 members of Groundwork’s Steering and 3 members of Groundwork’s NPC delegation to explain Groundwork’s perspective on the recent NPC decision to eliminate the all-member poll on the presidential election. We also want to clarify why there has been a strong reaction among Groundwork rank and file to this moment as well. We hope this piece will be sufficient for these goals.
Yesterday, DSA’s NPC discussed DSA’s Presidential Endorsement Committee’s proposal for an endorsement process for President in 2028. This proposal included an all-member poll, following multiple rounds of chapter and national deliberation, in line with what we did in March 2019 for the Bernie 2020 campaign. The NPC instead voted on and passed an amendment brought by NPC members from the Marxist Unity Group and Springs of Revolution to cut the all member poll and have so-called ‘chapter polls’ in its place. The argument made by these formations is that there will still be polls of member opinion, but we have instead empowered chapters to control and direct these polls, and they will instead be conducted according to the ‘standard processes’ of these chapters. Some may then ask- why has this change become such an issue for Groundwork?
When conducting experiments or practices in which one wants to acquire data, a critically important aspect of designing a methodology is the rigor of the process you apply. Rigor as a concept demands the elimination of as many differentiating variables as possible in procedures, to ensure an experiment can focus on the variables one does want to examine without conflicting noise from less desirable variables. Without rigor, it is substantially harder to draw strong conclusions from any process: the noise-to-signal ratio becomes too great.
One single mass member poll sent by National DSA controls several variables in how members are polled- the language used in the poll, the timing and frequency of the polls being deployed, and most notably, who is polled can all be consistent across the entire base.Whatever one’s position may be on the democratic character (or not) of polls, a single, large poll with all variables controlled has high utility because the testing conditions are constant. But 230 different polls spread across all our chapters completely confounds this. Chapters will have varying levels of follow through, communication and framing on the questions of the poll. Some chapters inevitably will not do a poll at all. And, most importantly, different chapters will poll different subsets of their members.
If there is one big national poll and we see the results of it, inevitably the discussion will center on what the results tell us about the membership’s opinion on the question. But if X number of chapters do a full member poll, Y number of chapters do an in person vote, Z number of chapters only poll in person those who are engaged in a specific working group, etc etc, the quantitative interpretation of the poll is innately contestable, even when aggregated. Discussion of the poll will not focus on what it says about members’ thoughts, because it cannot be clearly ascertained via this structure what members’ thoughts are in a clear and quantitative sense.
Instead, in the new system, assuming everyone has access to the results, different factions will gravitate to different explanations based on their ideological priors. GW and SMC will likely aggregate the vote totals and may (assuming they are majority in favor of AOC) call it a mandate. But other factions and members may insist the proper comparison is by number of chapters, not number of members. Some others may assert the in-person votes are more legitimate than the full membership votes. Others may say it’s wrong to quantify the results instead and that the best thing the NPC could do is examine the quality of for and against statements. And some may even just say the whole thing is impossible to interpret.
The point is, without a clear, consistent measure, the degree to which usable information can be extracted collapses, and everyone will retreat back to their ideological priors. The poll becomes another piece of noise by which to engage in the same usual lanes of DSA debate of the presidential, and by proxy AOC, question itself.
————
Many of those against the poll insist that this information is useless anyway, however. Many NPC members who voted against the all member poll claimed that AOC’s endorsement is inevitable, even while saying they themselves will likely vote against it, and suggest any concern for a removed poll is an overreaction given this.
This reasoning is unpersuasive to Groundwork (and we assume Socialist Majority Caucus, who also supports a member poll) for two reasons: firstly, the vote on having this poll is inevitably going to be read as a proxy vote on how people feel about AOC (particularly in the case of Bread and Roses, who didn’t come out in opposition to the poll they helped add to the PEC proposal until the day before the NPC meeting). Every caucus in DSA at this point seems to assume that an all member poll on the question of endorsing AOC would result in likely supermajority support. Given it is nonbinding and setting aside abstract commitments to being anti-poll (again odd as a primary concern given the body ultimately did pass a convoluted and confusing set of chapter level polls, voted for by the same caucuses who claim such an ideological commitment), the most obvious tactical reason a formation would oppose an all member poll is it obviously makes it politically difficult to vote against a presidential endorsement if the membership is seen to overwhelmingly support it (notably, this is precisely what DSA members have long accused the Working Families Party of doing in the 2020 election, when they did not release the vote totals for polled members and endorsed Elizabeth Warren instead of Bernie Sanders).
Secondly, contrary to these arguments, the outcomes of such a future vote are not inevitable. None of the formations opposed to a poll have announced that they plan to vote yes on an AOC endorsement. All of these formations will instead have internal deliberations at some point to decide this, deliberations the results of which we do not yet have. These deliberations will then decide how the NPC votes. Now, none of these internal caucus deliberations will in reality be shaped by the results of membership polling, but solely the attitudes and concerns of the members and leadership of those caucuses. That NPC vote will be further shaped by the question of who is in attendance at that vote- not an inconsequential concern, given the vote split on this NPC is frequently 14-13 and any one person’s absence can swing incredibly consequential votes. The idea that an AOC endorsement is an inevitability, rather than an extraordinarily close vote predicated on 1-2 caucus’ internal deliberations and who is able to attend the NPC meeting, is illusory.
Indeed, it speaks to a deeper issue- if the vote will inevitably be close on the NPC, while all factions involved believe that the membership if polled clearly would overwhelmingly support AOC’s endorsement, then there is simply a shared acknowledgment that the NPC is substantially out of step with the wider membership on this question. Marxist Unity Group NPC member Sid C W suggested this at the NPC meeting yesterday when he stated his fear of an overwhelming poll mandate being followed by an NPC vote was that it would cause a crisis of legitimacy for DSA if the NPC decided to rule the other way from the poll. The consequences of this are fairly substantial, particularly if the NPC does not endorse AOC, despite multiple members currently begrudgingly stating it is an inevitability. The ability of the national organization to engage our wider membership is a matter of how well those members feel their desires and political concerns are represented by the leadership. In a situation where the NPC and broader membership are completely misaligned on the 2028 election, the broader membership will instead look elsewhere for leadership- a massive issue for those of us who want, and believe this country NEEDS, a strong national organization that can organize millions of people to be Democratic Socialists.
———
A narrative has emerged recently that whereas Bernie 2020 was an obvious decision for the organization, AOC is a controversial presidential endorsement that DSA is unsure on. With respect to those making the contrast, Bernie 2020 was in fact considered highly controversial by caucus formations in DSA in late 2018-early 2019. Much ink was spilled about how it would split the org, that DSA had grown past Bernie after 2016, and that the safest choice would be to instead let individual members coordinate while DSA abstained due to Bernie’s own political failings. To anyone tuned in to this discourse, DSA appeared split down the middle on Bernie’s endorsement.
Then, the March 2019 Bernie poll (originally proposed by the nascent Bread and Roses caucus) was held, and the results released to all members. 13,324 people voted in the poll. 76% said Yes to Bernie and 24% said No- a +52 result for a Bernie endorsement. With this, all talk about the internal controversy fell apart. People who had been strongly opposed just months before started organizing canvasses for Bernie in their chapters, pulled along by the sheer power of mass support for the effort. The controversy evaporated and the entirety of DSA essentially united in the effort.
With the all member poll eliminated from this decision, the NPC has severely limited the power of creating that sort of unity in action via mass participation. Even if one is opposed to polling in principle, it is easy to see that support for a still not yet existent AOC’s campaign among DSA’s rank and file is inevitable. It will only increase if she announces, and DSA’s members will turn out in droves for her whether the national organization has moved or not. The question is, will that participation be mediated by a national body that can elevate our national platform, be in the campaign’s team as DSA, and do our utmost to make the campaign a DSA effort both in practice and to the public? Or will that energy instead be primarily mediated through those caucuses and chapters passionate enough about the AOC 2028 effort to engage, allowing growth and leadership on the ground for them but blunting our ability to exert ourselves as a national force with influence on the strategy and message of her campaign?
We worry that we have resigned ourselves to the latter. We hope for the purpose of national DSA becoming the socialist movement we want to see in this country that we are wrong. We will find out over the next year. In the meantime, we encourage DSA members all over the country to engage in good faith with whatever process comes out of the National organization, as well as organize for their chapters to endorse AOC when the time comes next year. This is not the end of this question of DSA taking on the presidency- it’s only the beginning. We are all in this together for the long haul, as one DSA.
