Skip to content

Strange alchemy

The party surrogate and socialist politics in DSA

Andy Sernatinger and Emma Wilde Botta critique a strategy popular among leaders of Democratic Socialists of America that places socialist politics inside the Democratic Party.

There’s a new consensus growing among leaders and caucuses of the Democratic Socialists of America (DSA): there will be no break from the Democratic Party. At least, not any time soon. Since 2016, the electoral debate in DSA has been about how to build an independent party. It was taken as a given that the Democratic Party is an obstacle, though the Bernie Sanders campaign opened up the possibility of tactical engagement with the Democratic Party as the socialist movement fought to move past it.

Starting in 2020, the terms of the debate have shifted to whether to break from the Democratic Party at all. Leading players who argued over this question in years past now all seem to agree with each other: breaking is bad. Prominent writers on the dirty break and the democratic road to socialism—Eric Blanc, Chris Maisano, and to a lesser extent Neal Meyer—have revised their position to one in favor of indefinite runs as Democrats, until such a time when “the contradictions” solve the problem for us. If you read social media, they agree there is little distinguishing their approach from that of groups they previously opposed in DSA.

The new common sense is centered on a concept called the “party surrogate”—a “party-within-a-party” model promoted or adopted in various forms by DSA caucuses such as Collective Power Network, Emerge, Red Star, and Socialist Majority. The “party surrogate” is used as a prop to argue against political independence from the capitalist parties. They use a lot of complicated formulations and arguments (ballot lines are state institutions and there is no Democratic Party) to end up saying something simple: socialists should just run as Democrats.

How, in the space of only a few years, did the aim move from using the momentum of a political moment to break free of the constraints that have hindered Left politics for nearly a century, to abdicating the project entirely? How is it that every major DSA caucus now agrees on questions that previously caused major divisions? What is the “party surrogate” that’s been touted and where does it fit in?

The dirty break, broken

DSA as we know it is the product of a particular political conjuncture starting in 2016: the Bernie Sanders campaign and the election of Donald Trump. For years before then, DSA had a stable membership of about 6,000 members —it wasn’t all that different from other socialist organizations that maintained essentially flat membership rolls, though it kept a larger paper membership.

The Sanders campaign served as a catalyst and DSA a potential receptacle, but there was still the problem of the Democratic Party. For people who would make up the bulk of the new DSA, the previous decade hadn’t instilled confidence in the Democrats. The Obama years left a competing sense of low faith in the Democratic Party with high urgency to defeat the growing Republican-allied far right. To move beyond the initial boost in DSA membership, there was a project to construct political theory that could address the concerns with the Democrats and create a bridge that would lead to supporting Sanders and other candidates when they would run in the Democratic Party.

This project had two wings: the dirty break, developed by Eric Blanc from 2015 on, and the party surrogate, first raised by Seth Ackerman in 2016. Both of these theoretical bridges—surrogate and break—have similar approaches: they acknowledge the need for working-class political representation; they reason that barriers to independence are too great (at least, right now);  then they propose a way of running campaigns in the Democratic Party, but not of the Democratic Party.

Both started with the desire to get the socialist left to support Sanders when he would run again in 2020, and worked backwards from there. Neither model could simply continue DSA’s historic position on realignment—there was too much history pointing to the failures of trying to reform the Democratic Party. These projects needed to advocate for running within the Democratic Party in spite of the history.

It’s the dirty break that defined the early sentiment in DSA, becoming codified at the 2019 national convention through the “Class Struggle Elections” resolution. Jacobin writers Meagan Day and Micah Uetricht incorporated dirty break theory into their book on Bernie Sanders, and DSA chapters hosted numerous events projecting dirty break politics.

So what happened? Less than a year after adopting Class Struggle Elections, the perspective shifted from open struggle in the Democrats and “agitat[ing] within the party, in full view of the party’s base…making it harder to ignore” (Day & Uetricht, 72) to camouflaging, or muting that conflict (Neal Meyer). Meyer writes, “In our campaigns, we plaster ‘Democrat for…’ prominently on posters and literature for our candidates — precisely because we know how powerful that symbol is for voters who despise Republicans.” Chris Maisano expresses this plainly: “the Democratic Party question…[is] settled, at least for now”.

The dirty break is fairly easy to understand, but its liability is that it had to demonstrate that it was preparing the ground for an actual break. The theory itself is entirely unclear about how this would happen but suggested that the Sanders campaign—despite his own opposition to a break, dirty or otherwise—was intensifying contradictions that would solve this problem without socialists having to really build independent politics. The truth was actually the inverse: dirty break theory depended on Sanders to contain the contradictions of its own formulations, and once Sanders was out of the picture the problems could be deferred no longer.

The Democratic Party organized to crush Sanders for a second time in 2020, and, rather than press the challenge at the Democratic National Convention, Sanders conceded and immediately endorsed Joe Biden. The anchor of the Sanders campaign was pulled up, and the ship of democratic socialism quickly drifted. Dirty break advocates like Blanc, Meyer, and Megan Svoboda campaigned for Joe Biden, contradicting dirty break rhetoric about “be[ing] content to withhold endorsement in the event no [class struggle] candidate is running” (Day & Uetricht, 129).

The emphasis that dirty break theory put on running in Democratic Party primaries trained DSA members that working within the Democrats was a necessary part of politics. These attitudes were hardened as leading DSA members took staff positions for Democratic politicians, adding material investment. The absence of any qualitatively different practice to build independence exposed the underlying inconsistency: “if what we’re doing is good, why do we need to break?” Without Sanders or a practice that built independence, the dirty break was broken. The political conjuncture that built DSA (Sanders and Trump) was suddenly over by the end of 2020. Articles by Maisano, Blanc, Meyer and others agreeing with their previous opponents on the party question betray this admission.

The forces in DSA that were more forthright about this reality benefitted. Collective Power Network (CPN) had been arguing against the dirty break since their formation in 2019, espousing a version of party surrogate theory, which was then taken up by other caucuses such as Emerge and Red Star. With the dirty break losing currency, CPN was bold enough to go beyond advocating for a “surrogate” and present “Breaking Bad”—a theory for the new conjuncture in the Biden years.

Time to party (surrogate)?

The “party surrogate” is quickly becoming popular among DSA leaders, but what is it? The term sounds highly technical and possibly related to pregnancy.

In 2016, Seth Ackerman sketched Blueprint for a New Party, his perspective on the party question. Ackerman put forward a proposal for a kind of party-within-a-party, an organized bloc that walked around the two-party system. He explained that this would be the “equivalent of a political party in…key respects: a membership-run organization with its own name, its own logo, its own identity and therefore its own platform, and its own ideology.” This organization would act like any other political party minus its own ballot line. It would develop and support candidates who reflect the political will of the organization and aim to generate its own working class base that pulls in Democrats, independents, and non-voters.

In Ackerman’s formulation, this party is “a national political organization that would have chapters at the state and local levels, a binding program, a leadership accountable to its members, and electoral candidates nominated at all levels throughout the country.” He’s silent about long-term goals, choosing instead to focus on the necessity of an independent base. Importantly, Ackerman expresses no preference about the ballot line: “The organization could choose to run in major- or minor-party primaries, as nonpartisan independents, or even, theoretically, on the organization’s own ballot line.”

In 2019, Dustin Guastella and Jared Abbott expanded on Ackerman’s ideas, with what they called the “party surrogate.” In A Socialist Party in Our Time?, Guastella and Abbott add two new elements. First, they offer a more detailed analysis of the problem, the core thesis being that “the state and party are now so intertwined that it is sometimes hard to find where government ends and parties begin”. This leads adherents to argue that there is no Democratic Party and that ballot lines are state structures.

Second, they flip the political issue, writing, “party realignment and exit are better understood as effects of a successful campaign…and less as strategies towards that end.” We’ll only break or realign the Democrats if we do the party (surrogate) thing well. (Note that here they do recognize the Democratic Party as an institution.) It’s only a hop, skip, and a jump from there to arguing that the party surrogate must be exclusively inside the Democratic Party. Guastella and Abbot write, “The party-surrogate would function similarly to a mass party but would make use of the Democratic Party ballot line.” This is different from Ackerman’s earlier view that the “party of a new type” would run on any ballot line it could, including its own.

In 2021, CPN articulated its own version of party surrogate theory. Brad Chester, himself a professional Democratic Party operative, steps in with Breaking Bad to provide the rationale for running candidates exclusively in the Democratic Party and extends the arguments to argue against a break. Chester argues that refusing to compete in Democratic primaries “establishes a permanent roadblock to elected power for socialists in America.” He argues that by the time a large enough constituency was built up to be able to break, that voting base would be powerful enough to become dominant in Democratic primaries. “At that point, it doesn’t matter what ballot line we use, we’d already be delivering all the material results for the working class.”

With each new iteration, the party surrogate becomes more adamant about socialists only running as Democrats. By the time it gets to Chester, anything other than the Democratic Party is attacked as “playthings of cynics who have opted out of mainstream politics, not a serious vehicle for organization.” Where Abbott and Guastella argued that a break is one possible result of a successful party surrogate, Chester dismisses breaking as entirely unnecessary. This is a big transformation in the span of five years.

The rules ain’t the problem

The party surrogate premise (create a party-within-a-party) quickly becomes complicated. It starts with the problem, the absence of a labor-based third party in the U.S., and explains that the reason third-party efforts have failed is because of the rules of the electoral system. If we had a more democratic system, a party would have already formed, but the U.S. is exceptionally bad, so it’s too difficult.

From there, surrogate proponents argue that the U.S. doesn’t have a true party system and party ballot lines are in fact “state institutions,” which makes them things to struggle over. They extend this further to say that the Democrats are not a “party,” and that the Sanders campaign showed, in fact, that it is possible to run candidates of any sort on any ballot line. Therefore, socialists should run exclusively as Democrats.

There are a number of problems with the practical, historical, and conceptual layout of the argument. Because there’s such a direct progression of arguments and quick conclusions that all work off each other, let’s start at the end and work our way backwards.

Primarying our way to second fiddle

The culmination of party surrogate (and dirty break) theory is the argument that socialists should contest elections in the Democratic Party, particularly through partisan primaries.

If ballot lines are state institutions or hollow bodies to be contested, why confine yourself exclusively to one? Ackerman was agnostic about which party to contest in, but the model has since resolved to be exclusively in the Democratic Party. (To our knowledge, no DSA socialists have contested Republican primaries.) This suggests a corresponding shift in the idea of how a party surrogate should work—rather than building up its own base with its own identity, the surrogate is dependent on an assumed base around the Democratic Party. Chester makes this clear when he writes that a mass base would “necessarily require a majority of Democratic voters and be powerful enough to dominate in Democratic primaries.” The assumption is that Democrats have the base socialists want.

In reality the base of the Democratic Party is shifting rapidly, where “Democrats are replacing Republicans as the preferred party of the very wealthy.” The richest U.S. counties are now swinging Democrat. Historian Matt Karp explains for Jacobin,

In its leadership and cultural style, the Democratic Party has already become a party of the professional class, with little interest in challenging the plutocracy and little capacity to deliver material benefits to workers of any race. If the shareholders of Amazon, Pfizer, and Goldman Sachs all understand this, why shouldn’t the iron miners of northern Minnesota?

The working class base of the Democratic Party is leaving it’s orbit. Nearly half of union members now regularly vote Republican. It’s not just the “white working class”—in 2020 Trump increased his share with Black and Latinx voters. Voters who make less than $50,000 (disproportionately workers of color) made up 47% of the Democratic electorate in 2000 but dropped to 37% in 2020. Kim Moody explains,

What has actually occurred in U.S. politics over the past few decades…is more of a stealth realignment in which voters of different social classes have switched from one party to another. Two trends in particular affect the Democratic Party and its electoral prospects: the relative decline in the working-class and union household votes that began long ago, on the one hand, and the Democrats’ more recent increased dependence on well-to-do and wealthy urban and suburban voters, on the other.

It’s not as though there are no working-class Democratic voters, but the class character of the party has shifted. Though the Democratic Party may be the party of labor leaders, it is less and less the party of laborers.

This has practical implications for socialists. Elections are skewed in class terms, where voter participation is directly linked to income. This is even more exaggerated in primary elections, since working-class voters predominately show up to general elections, if they vote at all. Primaries are dominated by the middle class and the wealthy.

Source: Brookings Institute

For socialists who contest these elections and who are set on “winning,” they necessarily have to tailor a message to the classes in the electorate who show up to vote in a primary. If you lose—and statistically if there’s an incumbent you will lose—you’ve ceded the ability to present  candidates in a general election who might reach the working-class voters you’re aiming to attract in the first place. This ends up being a conservatizing strategy; you have to relate to a different class in the primary in order to make it to the general.

The Democratic Party exists, actually

For party surrogate theory to work, you have to accept that the Democratic Party is an empty vessel. The argument is that there actually isn’t a Democratic Party, just a ballot line. But if there’s no Democratic Party, was there ever one? Was there a Democratic Party in Chicago 1968? Was there a Democratic Party during the New Deal? Was there a Democratic Party when it formed in 1828? If at any point there was a Democratic Party, where has it gone? If it did exist and now it doesn’t, did it abolish itself? It seems silly, but it’s an important question in understanding the development of the modern Democratic Party.

Abbott and Guastella step on this accidentally when they write about the decline of party machines, but they take this to mean that the Democratic Party was “hollowed out” when local machines declined. Neither Ackerman nor Abbott and Guastella mention the Democratic Leadership Council (DLC), the Democratic Congressional Campaign Committee, the Democratic Governors’ Association, or any other standing party organization. It’s easy to understand why they end up thinking that the party is an empty shell when they’re silent on these major institutions that strategize, develop policy, rear candidates, and run the party.

In the 1980s, the Clintons and the DLC started to remake the Democratic Party. Democrats of the post-war years were dependent on local party machines and courthouse gangs, which bankrolled the national party and could make demands on state government and presidential candidates. The DLC viewed machines as insufficiently willing to serve capital and too susceptible to public pressure. They sought to remake the party by centralizing its functions, particularly fundraising, and creating a professional core that would follow directives. The decline of machines didn’t end the party — it shifted power towards a centralized apparatus of permanent professional staff, politicians, and funders. This power compact was able to dominate and direct the Democratic Party towards welfare reform, prison expansion, and neoliberalism.

This doesn’t mean that there was no “Democratic Party” when machines and courthouse gangs faded from view. Rather, the Democratic Party mechanisms used in the period of high neoliberalism were different from those used in the post-war era. The contention that there is no Democratic Party assumes that all political parties are mass-democratic parties. They’re not. Bourgeois parties aren’t interested in democratic governance—they have no use for such things. The core of the Democratic Party are the staff, officeholders, and funders, with a secondary layer of feeder groups that stump for the party consistently, including organized labor, feminist, civil rights, and environmental organizations.

The Democrats adapt to new developments by allowing for a porous concept of what is in the party bounds, but only at the edges. As a bourgeois party, it allows for competition between potential representatives as to who can best manage the attitudes of the shifting electorate while fulfilling the mandate of capital. The Democratic Party as an institution has to be able to adapt to new developments and incorporate them in some mutated form in order to maintain viability, or else it would have gone the way of the dinosaur.

There are limits to how far adaptation will go. The most obvious is how the Democratic Party selects their presidential candidate. Some states have “open primaries,” where anyone can vote regardless of party registration, while in others only registered Democrats can participate. What they actually vote for are delegates to select a candidate at the Democratic National Convention, which remains a private party affair. If the “soft” mechanisms that manufacture results fail, by beginning with caucuses and primaries in conservative states to signal electability to subsequent voters, the arcane convention system remains with its superdelegates selected by the permanent party bureaucracy. Guastella’s analysis embarrassingly led him to argue in early 2020 that “After Nevada, It’s Bernie’s Party Now”—he had to eat those words not a week later. The core of the Democratic Party is highly centralized and intervenes regularly to maintain its interests. The permeability at the edges of the “ballot line” ends up being an essential function of the party’s hegemony but this doesn’t shift the general policy or direction of the party nor weaken it’s control.

Good kid, bad system

All of these complicated formulations about a party surrogate rest on a simple idea: the rules are too harsh in the U.S. and it’s just too hard to form an independent party. The arguments that follow, about ballot lines, primaries, legitimacy, and whatever else, are all dependent on what turns out to be a basic impulse. From there, they prune history and use an upside-down concept of how political parties are formed.

What all of these thinkers have in common is that they argue for a new American exceptionalism—it’s not the Protestant Ethic, the absence of feudalism, or cheap land that prevented a labor party in their account. It’s that the U.S. is uniquely bad with its electoral rules.

Is it actually different though? The U.S. is not the only country with a single-district, first-past-the-post, winner-take-all system. India, Australia, Canada, and the United Kingdom. have similar configurations but have more than two major political parties. This isn’t to say the U.S. doesn’t have some distinct features, like all countries do, but this system alone doesn’t automatically lead to a two-party system that you just have to accept. Rather than being a primarily structural problem, it’s a historical and political problem. Take the observation by sociologist Robin Archer,

Before 1900 no European country used proportional representation for national elections, and no large European country used it before the end of the First World War. Thus, these obstacles, far from being unique to the United States, were actually the norm.

The rules that some now think allow for multi-party democracy in other countries were actually the result of parties that struggled for representation in a system not entirely unlike what we’re still dealing with in the U.S today. The difference is that a labor-based party in the U.S. never became strong enough to force these changes from the representative state.

Reading party surrogate theory, you get the impression that the system has always produced the same results and so the only responsible thing to do is accept it and adapt. The history they cite is surprisingly light: James Madison didn’t want political factions, and then, Ackerman writes, in the 1880s and 1890s “a series of state-level legislative reforms that permanently transformed the American political system, creating the electoral machinery we have today.”

They say nothing about the Socialist Party of America (SP), founded roughly ten years after the period Ackerman mentions. The SP had hundreds of elected officials in local governments and state legislatures, and even elected some to Congress (though the federal government refused to seat them). It wasn’t bad rules that spelled the demise of the Socialist Party; it was a combination of state suppression, imprisoning or deporting party leaders, and self-destruction by the SP right-wing, which refused to concede leadership after losing control of the executive board. It’s actually in response to the success of the SP that reforms were enacted in most local elections after the SP ceased to be a unified force, changing local elections from partisan (party-line) to non-partisan races with general primaries.

The SP is significant as a case of a political party that was successful while operating in the same first-past-the-post, single district system of the U.S. It only stopped working because the state tried to crush it, and then the rightwing SP leadership expelled the majority of the membership. Building a labor party was common sense until the goal was abandoned with the invention of the Popular Front and the New Deal Coalition in the late 1930s.

Still other attempts were made at forming a party, but the primary obstacle was that organized labor was allied to the Democratic Party, not that the electoral rules were exceptionally bad. Better rules could have helped, but they wouldn’t solve the problem of the base. (Civil rights, feminist, and environmental groups would also ally themselves to the Democrats, contributing to the problem.) The Popular Front, realignment, and similar notions have starved the project of a labor party of the people who would work hardest to build it: socialists and communists.

With both party surrogate and dirty break, the process of party building is inverted. Their concept of party formation is dependent on national politicians (Sanders, Alexandria Ocasio-Cortez), whose campaigns and offices then trickle down to state and local efforts to fill a political party from the top down. It starts with national politics, which then makes state and local races possible through the same “party surrogate”.

This is extrapolating from the Sanders campaign and it assumes that Sanders and the Squad are in fact committed to a party surrogate (or a dirty break)—they are not. In reality, the efforts of an organization like DSA follow these candidates rather than direct them. We don’t tell Sanders or AOC what their positions should be; rather, they tell us what ours should be. That doesn’t resemble anything like a democratic political party.

Examples like Jeremy Corbyn in the U.K. or Jean-Luc Mélenchon in France suggest that reliance on a unifying national figure hasn’t even been effective even with parties that are established or where the rules are more favorable. It creates a disembodied pool of supporters that don’t have strong enough social bonds or organizational infrastructure to leverage their numbers. (The U.K. Labour Party swelled with members in 2019, but these were almost entirely at-large members rather than members of neighborhood or union affiliates.)

The formation of political parties is dependent on creating organizations that then band together to grow their scale. The Farmer-Labor Party strategy of the 1920’s and 1930’s was the development of state parties that would then be able to link together to create a national party. Again, these efforts were frustrated more by alliances with Democrats and the sectarianism of the Communist Party than by structural difficulties. The ability to run national electoral campaigns related to democratic membership organization requires grassroots organizations strong enough to support candidates but also to assert the will of their membership rather than get swept along by the politician’s campaign.

Lead into gold

The agreement between DSA leaders on electoral strategy flows from a shared belief that the problem of political parties in the U.S. is a structural rather than a political problem, which is insurmountable. They depend on an idea of the Democratic Party as decentralized and open, ignoring the centralized party organization made up of professional staff, funders, and party officials. This party center does in fact intervene and assert power, as it did in both of Sanders’ primary campaigns. The conclusion drawn by party surrogate and dirty break advocates is to adapt to “bad rules” by running as Democrats indefinitely. It’s a strange alchemy that claims to have taken the lead of the Democratic Party and turned it into socialist gold.

Dirty break and party surrogate strategy share a trickle-down conception of party-building. This idea reflects the hopes generated in the Sanders moment that a national candidate could build the party from the top and everything else would fill itself in. In the course of pursuing these electoral campaigns, many activists took paid positions as campaigners, staff, surrogates, or just deepened their relationships with Democratic partisans. This helps explain how a hot war between DSA factions only a few years ago has since turned into political alignment.

At the end of the day, neither the dirty break nor party surrogate describe DSA’s electoral work up to this point. Though Abbott and Guastella state in their 2019 article that “DSA already carries out its electoral work in a manner similar to the type of party-surrogate we’re advocating,” they left unexplained why that’s the case. There hasn’t been any party-within-a-party infrastructure built. Very few candidates are brought up from the ranks of DSA to run campaigns as an expression of the organization’s politics, and even fewer campaigns are initiated within chapters through deliberation and debate. Instead, most DSA-endorsed candidates come to the organization for endorsement, often with predetermined platforms and no expectation of being held to account.

At the heart of both models is an implicit assumption about the path to socialism: it will be elected in (“the democratic road”). The success and strength of our movement is measured by electoral wins. Reforms that benefit the working class are seen as a product of electoral organizing rather than class forces.

The fundamental problem with this strategic vision is that it leaves out the agency of the very people with the power to build socialism. It assigns a largely passive role to workers and oppressed people, instead of seeing their collective strength and creative potential as the foundation for a socialist transformation. It indicates a cynicism about the ability of the working class to change social conditions and political structures. That disbelief underlies the constant revision of the party surrogate idea to attach itself more firmly to the Democratic Party.

If you don’t think you can even change electoral rules, how do you think you’re going to mount a challenge to capitalism? That ends up taking someone like Chester down the road of saying that it all comes down to the gains you make on behalf of the working class. Class consciousness and the organization of the working class into its own political party for the purpose of contesting for power leaves the equation entirely.

We’ve focused on critiquing a strategy that places socialist politics squarely inside the Democratic Party. The positive case for an independent party and a sketch to actualize it can’t fit here but will be taken up in a future article. These are questions that are asking us what we think socialism is—are the different roads going to the same place, or to an entirely different project?

An earlier version of this article listed Viewpoint among groups favorable to the party surrogate.

We want to hear what you think. Contact us at
And if you've enjoyed what you've read, please consider donating to support our work:


Andy Sernatinger & Emma Wilde Botta View All

Andrew Sernatinger is a labor activist and member of DSA in Madison, Wisconsin. He is a member of the Tempest Collective and has written for New Politics, International Viewpoint and Jacobin. Emma Wilde Botta is a member of the Democratic Socialists of America and the Tempest Collective.