Earlier this month, the Democratic Socialists of America (DSA) held its 2019 National Convention. I wrote a piece for Jacobin giving an overview of the major decisions that came out of the convention, but there’s a larger conversation to be had about the experience at the convention in Atlanta.
Immediately following the convention, several accounts addressed the problems of convention: the numerous points of privilege, posturing and procedural maneuvering. Allison Geroi sums up the sentiment in her piece Democracy Beyond the Debate Floor when she writes, “DSA’s 2019 convention was chaotic and often divisive. All major proposals inspired heated emotional debates, and every day featured equally contentious procedural motions.” To DSA’s left, Tatiana Cozzarelli writes for the orthodox Trotskyist Left Voice that the “procedural problems of the DSA convention was an expression of its reformist orientation.” The New York Times’ Marc Tracy more charitably explains, “But if most delegates would have admitted that the emphasis on procedure lapsed into self-parody, some argued that it stemmed from an authentic desire to accomplish things.”
Clearly things did not go as smooth as they should have; nearly the entire first day was taken up with procedural motions to decide on the rules of convention, rearrange the agenda, extend and un-extend times and disruptive points of privilege. The convention continued to lag behind schedule, and as the body approached its final time limit numerous motions were packaged together to try and pass them as quickly as possible, without the debate the motions deserved. How do we make sense of this complicated, often frustrating experience? Is our hoaxster comrade Archie Carter correct that “DSA is Doomed”? Is it, as Left Voice argues, a symptom of reformism? Was this just factional bickering that held the convention hostage?
In what follows, I’d like to examine the convention process and offer some thoughts about why there were problems with the convention, interrogate how serious those problems are and offer some thoughts on the democratic tasks for DSA.
Lead Up to Convention
Let’s start by revisiting the timeline of convention. Most accounts have focused on the weekend of convention itself, but we should start by looking at how DSA prepared in order to understand how the months leading up to it set the stage for the convention.[i]
In mid-December 2018, DSA National Director Maria Svart announced the National Convention would be held in Atlanta, Georgia, from August 2nd through 4th. To prepare for this, the national organization would organize a series of pre-convention conferences to be held from February to April of 2019 across the United States. Nine regional pre-conventions were held in the late winter and early spring with a program designed to discuss the priority areas of DSA’s work since 2017 (elections, labor, Medicare for All) and Bernie Sanders and the 2020 election. The pre-conventions also had nuts and bolts workshops on how to write proposals for convention, how to use Robert’s Rules of Order and an explanation of the logistics of convention.
After the last of the pre-conventions in April, most activity was centered on chapter elections of their respective delegations to convention and on the formulation of proposals for the convention, though both of those tasks were seemingly isolated; chapters turned inward to decide who they would send to Atlanta, and as I’ve written about previously most of the proposals were written by caucuses. The compendium of eighty-five (85) resolutions and twenty-eight (28) constitution/bylaw amendments were released at the end of June, which left a little under five weeks to read over 260 pages of proposals and 138 pages of candidate statements. Immediately following the release of proposals, there were approximately two weeks to submit amendments to the proposals, which led to a kind of mad scramble to draft amendments.
Delegates were then asked to vote in two straw polls in advance of the convention. The first straw poll on July 15th, 2019, was to decide on a “consent agenda,” a package of proposals with enough approval whereby we would agree to forgo debate and vote on them as a bundle in order to save time at the convention. A second straw poll sent a week later then asked delegates to vote on which proposals they wanted to prioritize for debate on the convention floor, as there were far too many to be able to address them all. (We’ll return to these in more detail below.)
It might not be immediately evident, but the regional pre-conventions and the Atlanta convention were in effect two separate events with little relation to one another. Because the first round of proposals hadn’t been submitted until June, two months after the last pre-convention, DSA did not actually use the pre-conventions to begin the conversations about proposals for convention that could have led to clarification, revision, or just better debate. The pre-conventions instead worked as regional conferences that were useful in building connections between chapters but didn’t prepare DSA for what would happen in Atlanta. Many delegates had only a passing familiarity with the resolutions and amendments and the body spent precious time in Atlanta introducing and explaining the proposals, clarifying their intent and effect, rather than debating their political impacts.
What we saw in Atlanta was the result of a compressed convention: the basic tasks of becoming familiar with proposals, beginning debate, clarifying and refining, were all waived in the pre-convention period. Negotiations that should have begun at the pre-conventions about the priority of the schedule, norms for debate and specific language that required change all got lumped into the convention weekend in Atlanta, putting tremendous pressure on delegates to get through everything on the agenda, ensure that their positions were heard and respected, and fight to correct problems in the process that they didn’t think they had a way to resolve before the convention. No organization could meaningfully work through all these needs in the span of three days, and many of the problems with convention stem from these constraints.
The Role of Caucuses
Another popular reflection from convention was that it was simply the result of factional positioning. In this narrative, the proceduralism and decisions made were basically a slugfest between organized national tendencies. Indeed, I wrote pieces explaining the influence of caucuses in producing the resolutions and amendments for convention, and in deciding who would stand for election to the NPC. But viewing the entire convention as just a feud between caucuses oversimplifies.
First, its worth noting that there was generally a high level of political agreement, which crossed factional lines. The convention debated approximately thirty-six (36) proposals, passing twenty-five (25) of those and referring three (3) to committee; when we consider only the political resolutions, nineteen (19) of twenty-one (21) were passed (91%). Most of these resolutions passed with a supermajority. Three passed narrowly (Anti-Fascist Working Group, package of Cuba Solidarity, Decolonization and BDS Working Group, and Rank and File Strategy), each with their own logic.
The proposal on anti-fascism was not so much a debate about whether DSA should declare itself anti-fascist, but on the tactical decision to form a national working group within the organization. Antifa has developed more as a decentralized movement rather than an organization, so discussing a national working group was about assessing the effectiveness of Antifa and whether it should be more coordinated. To my knowledge, this was the first discussion of an approach to antifascism in DSA and the split vote reflected our uncertainty.
The bundle of proposals on Decolonization, Cuba Solidarity and a BDS Working Group caught many by surprise, and many delegates voted against the resolution out of frustration with the procedural move to immediately place the resolutions at the top of the agenda; because they knew very little about the politics of the groups DSA would attach itself to with regards to Cuba; or because two of the proposals were not on the schedule and delegates were simply unfamiliar with them and voting them down was the safer bet.
The Rank and File Strategy (RFS) proposal was the most fraught, and this could be more factionally aligned than other proposals in that it was advanced by Bread and Roses (B&R). However, there are three important qualifiers: 1) this resolution immediately followed another “Towards a Clear, Multifaceted Strategy for Labor” (advanced by Collective Power Network), which led some to believe that the proposal was redundant; 2) the RFS proposal asked DSA to take a more nuanced position on how to build the labor movement and orient towards the labor bureaucracy, which was beyond the lived experience of many delegates – given the historically low levels of union organization in the United States, it is not surprising that many delegates are unfamiliar with of labor politics; 3) whereas most of the factional alignments had been along organizational lines (coordination vs decentralization), this vote did not follow that pattern. Bread and Roses advanced the RFS proposal, which was endorsed by the Libertarian Socialist Caucus (LSC) (who are normally antagonistic towards positions taken by B&R); Collective Power Network (CPN) opposed the resolution, where they had typically been allied with B&R on organizational questions. Socialist Majority (SMC) split their votes, with some members supporting and others opposing.
This suggests that while caucuses played a role, their alignment was not static but malleable/flexible once they departed organizational questions. That isn’t to say that caucuses played no role, but it would be wrong to broad brush the convention simply as LSC-Build vs. B&R/SMC/CPN.
The Straw Polls and the Organization Question
A major question leading up to the convention was about what would happen to the national organization. As DSA has grown dramatically, two competing visions of the organization were put forward: one for a well-funded, national organization that could better coordinate the activity of members; another that wanted to redistribute resources and function as a network of locals. Prior to the convention, it was difficult to know how the membership felt about these issues. Without some objective metric, the only way of gauging the debate was through the arguments put out on the internet (Twitter, social media, DSA discussion board and occasionally written articles). Going into the convention, there was anxiety about this unknown element: where would DSA fall on the organization question? It was interesting then to review the straw poll data for the consent agenda and convention schedule, because they gave us a way to measure the opinions in DSA.
The poll on the consent agenda gave us a first look at how delegates might position themselves. 605 delegates voted in this straw poll, roughly two-thirds of the delegates. The data needs some interpretation because there are multiple reasons why a person would decide not to vote for putting something in the consent agenda (disagreement with the resolution, agreement but desire to have it debated, lack of understanding but general support), but we can be confident that if a person voted FOR something to be in the consent agenda that they were affirming the proposal politically.
The results of the straw poll showed that none of the constitutional amendments were sufficiently popular to make it into the consent agenda, and in fact only one amendment (CB 32 Assert Our Commitment to Ecosocialism) reached over 50% approval; the average was 30% approval. The resolutions had a much wider spread, with the lowest support for R27 Proposed Priorities for DSA at 4.8% and the highest support for R6 Orienting Towards Latinx Communities at 88%. The resolutions that made it into the consent agenda commonly focused on immigration, ecosocialism/fighting the climate crisis, and gender justice. These reflected the collective anger and anxiety about major national issues in this political moment: ICE raids, concentration camps, abortion bans and looming climate chaos. This first poll seemingly confirms my thesis that there is political agreement in DSA but less unity on organizational questions.
The second poll on what should make it to the convention floor had a similar number of participants (676). The system for deciding what would make it into the agenda attempted to create a “weighted” system, which allowed for a mix of broad and deep support for a proposal. Delegates voted for their Top 20 resolutions they wanted to debate at convention, and then also for a Top 3 that were most important to them, which received a greater “weight.” The combination of rankings was then calculated to decide what would go to the convention floor. Proposals that made it into the agenda had to satisfy one of three conditions: very broad support (many people voting in their Top 20), very deep support (many people ranking in Top 3), or a high enough mix of the two.
Of the top thirty-nine (39) scoring proposals that went to the floor for debate, the average Top 20 score was 239; the median Top 3 score was only 15 (mean of 45). A delegate could vote for a proposal in both their Top 20 and their Top 3, and the average ratio of Top 3 per Top 20 (where a delegate ranked a proposal both in their Top 20 and Top 3) was 19% (median: 9%). While the metrics look technical, what’s important is that it gives us an impression of what kinds of support delegates had for the proposals.
The top scoring proposal, R34 Ecosocialist Green New Deal had far and away the most Top 20 Votes (555) and demonstrably wide support; the next highest Top 20 value was for R33 Invest in Political Education (364). The second highest scoring proposal, CB2 Require that National Pays Stipends to the Locals (“Pass the Hat”), was only ranked in the Top 20 by 38% of delegates (258), but of the delegates who voted for it, it had fanatical support: 211 of the 258 who voted to put it in the Top 20 also put it in their Top 3 (82%). This is particularly noteworthy when you consider that the average frequency delegates ranked both Top 3 and Top 20 was only 19%. Calculating a mix of broad and deep support was useful in bringing a conversation to the fore that was passionately felt by a strong minority of members, but when it came to the vote it carried roughly the same amount of support on the convention floor; Pass the Hat failed, seemingly without even a simple majority.
The data also suggest that while the charge of being a highly disciplined faction has been levied against Bread & Roses, Build was actually the tendency that whipped their votes the most; at 82%, I cannot see another explanation other than a deliberate, coordinated push to secure support for that proposal. To further support that notion, Build’s other proposal, R60 Reassert Our Commitment to Training, had the second highest alignment of Top 3 per Top 20, at 62%.
The two polls give us different insights into DSA. The first poll on the consent agenda measured the overall attitude expressed by the delegates and showed our opinions on political questions. The second poll gives us different information, and arguably it showed us the priorities of the organized tendencies/caucuses. The ten highest scoring proposals (not including ecosocialism) were all sponsored by national tendencies and had higher than average Top 3 per Top 20 support at 30% or above (average of 47%), carrying the highest share of weighted Top 3 votes. Every other proposal that made it to convention did so only by having broad support (Top 20 rankings).
Does this run counter to my supposition that the convention was not as factional as otherwise stated? Yes and no. There is evidence to show that the agenda was set in large part by the national formations, but the votes on resolutions themselves were not simple reflections of caucus power.
Looking at this data we also could have predicted much of how the convention decisions landed. The straw poll showed that there was not broad support for any of the constitutional amendments, and in turn almost none of them were passed. The two amendments that did pass (close amendment loophole, remove gendered language) did not fundamentally change DSA’s structure but brought the organization in line with where the membership believed it should have been. While there was a very vocal group advocating for more localist approaches, the data showed them to be an intense minority but a minority all the same.
Three groups (SMC, CPN and LSC) proposed three separate models of regional organization. Each of them scored relatively well in the second straw poll, but each of them in turn failed, either by being referred to the NPC (SMC, CPN) or killed outright (LSC). There doesn’t appear to be a single determining factor for why the convention did not approve the regional proposals so we have to take a wider view.
- The proposals were not well understood by most delegates. Other than CPN, there wasn’t a very public effort to educate about regional organizational forms.
- They competed against each other rather than present a united model.
- They were highly prefigurative and did not reflect the experience of most members.
- They were placed at the end of the agenda and were pressed for time and, after a process-heavy weekend, there might have been a conceptual push back against the models as more bureaucracy.
Whatever the case, when it comes to restructuring the organization there wasn’t enough momentum to carry them forward.
Conclusion: We Need a Theory of Democracy
Where does this leave of post-convention? Three things stand out.
First, the convention missed its opportunity to have full democratic debate in the construction of the preconvention conferences. Typically, a preconvention period is an extended period that allows the membership to raise political questions, advance perspectives, review resolutions and debate them over a longer timeframe. When the body finally meets, they do so with a functional understanding of the issues at play and the votes can then reflect the culmination of the debate. Because the preconventions all took place before resolutions were available, members weren’t able to use those conferences to prepare for the convention in Atlanta. We should be forgiving of the convention, since 2017 was essentially the founding convention of our DSA and 2019 was in turn the first convention to assess the organization we are building together. There are bound to be problems, and now we have collective experience to inform future gatherings.
Second, DSA needs to address caucusing and organized tendencies. Clearly, caucuses are playing an increasingly large role in the life of the organization, at least at the national level. This is to be expected in a growing organization with thousands of members and shouldn’t be viewed as innately negative. Especially in a multi-tendency organization, political perspectives can be reduced to the lowest common denominator. For members who elect to join them, caucuses can provide greater political coherence and training without demanding that the rest of the organization submit to these as norms. This supports the “big tent” character of DSA and is common in socialist parties around the world.
Yet one of the holdovers of old DSA is that the structure was not designed with caucusing in mind; if anything, old DSA wanted to discourage caucuses and never formalized any rules related to members organizing internally. The NPC candidates, for example, were presented as a collection of individuals with no relation to one another; resolutions were also presented as by individual authors. In a DSA where there are organized groups with standing perspectives, this becomes important information for members to be able to weigh in the decision-making process. DSA will need to consider how to navigate standing tendencies alongside those who do not wish to affiliate with a group so that there isn’t a major imbalance in the ability to participate in the organization.
Finally, DSA needs a theory of democracy beyond simply voting. Socialism is at its core the extension of democracy into every facet of life. In capitalism, we do not live in a democracy, and so we must experiment and learn to develop democratic practices. Insofar as DSA strives to be a democratic organization, we need to have a better concept of what democracy means. The convention performed admirably given its limitations but certainly left much to be desired. What were the democratic deficits? For one, democracy requires that information is dispersed. We cannot make meaningful decisions without access to the information that informs our choices. But it doesn’t work to flood the membership with raw data; we also must have a grasp of the constraints placed on us by capitalism and our social positions to understand the varying levels of participation. As working people in a voluntary organization, the structures and norms developed cannot assume an infinite amount of time – in order to embrace the differing needs, DSA needs to offer multiple levels of engagement.
Democracy will require a commitment to educating ourselves and each other about the issues; it is not enough to “win the vote” if we did not move anyone in the process. The biggest casualty of the convention was the debate: DSA voted up important positions prioritizing immigration, abortion access and sex workers’ rights, but forewent the debates that would have planted these struggles more firmly among the membership. In the proposals on organization, DSA maintained its structure but the debate is far from over. In order to have a strong, dynamic organization we will have to consider how to accommodate the minority positions in order to maintain our unity for the struggles to come.
There is no magic formula for how to get it right. Democracy is a moving target, and as we grow and change the requirements for democracy will shift as well. It is a grand experiment, and we are sure to get much of it wrong, but we are up to the task.
 This figure does not count the Consent Agenda, which was voted up but not debated, nor does it consider “amendments” to be proposals in themselves; it also individually counts the resolutions voted up in as packaged items (three housing resolutions, three international resolutions and three criminal justice reform resolutions).
 Note: Delegates had to fill our their ballots completely for both Top 20 and Top 3 to be accepted. If 676 delegates votes, there were a total of 13,520 votes allocated in the Top 20 and 2028 in the Top 3.
 The former stands out for political reasons: CB 29 Equity for Amendments was written in response to a political maneuver to lower the threshold of votes needed to pass CB 2 “Pass the Hat.” The intervention of an author of “Pass the Hat” in favor of closing this loophole certainly helped put CB 29 over the edge. However, in substance all this did was remove an arcane rule that most had come to view as illegitimate.
- [i] 12/14/18 – Maria Svart announces of National Convention in Atlanta, George August 2nd to 4th.
- 12/21/18 – Announcement of regional preconvention conferences.
- 1/4/19 – Spring caucus announces formation
- 2/2/19 — Socialist Majority announces formation
- 2/9/19 – First Regional Preconvention Conferences held
- 2/19/19 – Bernie Sanders announces campaign for 2020 Presidency
- 2/22/19 – Collective Power Network announces formation
- 2/27/19 – National Convention Planning Committee announces it is accepting volunteers for sub-committees.
- 3/1/19 – 5/10/19 – Nominations for At Large delegates opened
- 3/5/19 National advisory poll to endorse Bernie Sanders for President sent to membership
- 3/12/19 Sanders poll closes: 76% YES to 24% NO (13,324 members voted)
- 3/21/19 DSA’s National Political Committee votes to endorse Bernie Sanders for President. DSA national campaign plan and independent expenditure for Bernie Sanders put into action. (Video here)
- 4/2/19 – Submissions opened for resolutions for convention.
- Nominations opened for NPC candidates
- Chapter delegate elections opened
- 4/11/19 – Spring caucus reforms as Bread & Roses caucus
- 4/14/19 – Last of Regional Preconvention conferences held.
- 5/1/19 – Convention Committee releases preliminary report
- 5/3/19 – Maria Svart announces Democratic Socialist Labor Commission (DSLC) endorsement of Labor for Bernie.
- Members of Subcommittees announced:
- 5/12/19 to 6/2/19 – Elections for at-large delegates held
- 6/2/19 –
- Resolutions and constitution/bylaw submissions period closed.
- National at-large delegate elections closed.
- 6/16/19 – NPC nomination period closes
- 6/18/19 – Local delegate elections deadline
- 6/27/19 – Compendium of resolutions and constitutional/bylaw amendments released
- Amendments period opened for resolutions and constitution/bylaw changes
- 6/30/19 – NPC candidates list and questionnaire released
- 7/15/19 – Straw poll for consent agenda released
- 7/16/19 – Deadline for amendments to resolutions and constitutional/bylaw changes
- 7/23/19 – Second straw poll released to decide convention agenda
- Deadline for second straw poll on convention agenda
- Convention program/agenda released
- 8/2/19 to 8/4/19 – Atlanta National Convention
This is a great analysis, and it captures my frustration with the rush prior to convention with particular accuracy. I do have one small correction: you write that “Bread and Roses advanced the RFS proposal, which was endorsed by the Libertarian Socialist Caucus (LSC) (who are normally antagonistic towards positions taken by B&R).” Each of these three elements is true on its own, but presenting them all together paints a slightly different picture from what happened. B&R advanced R32 for the convention, and R32 includes language that the Democratic Socialist Labor Commission should “focus on pursuing [RFS].” LSC did generally endorse RFS back in June 2019, but did not endorse B&R’s R32 for convention. While LSC did not write any statement of non-endorsement for any resolution, I would look to the reasoning in the LSC endorsement of RFS: “organizing the workplace requires multiple different approaches.” In terms of labor resolutions, LSC did endorse R67, Organizing the Unorganized (passed), and R68, A New Operation Dixie (referred to NPC by default). In the spirit of full disclosure, I caucus with LSC.