In an insightful essay at the New Republic, Evgeny Morozov raises a powerful critique of what he describes as “Internet-centrism” in Steven Johnson’s 2012 book Future Perfect. Morozov identifies in Johnson’s book a strain of popular rhetoric which holds that the Internet is a model of decentralization, horizontalism, and leaderlessness, and that the world would be improved by applying these features to other domains.

While most of the essay is spot-on, I’d like to complicate and dig a little more into the role of leadership in horizontal decision making and organizing within activist movements, one of the subjects of Morozov’s critique.

Morozov cites an insider from the Occupy movement, who describes how some members would stop going to the long, unproductive General Assembly meetings, but would nevertheless engage in sly, under-the-radar organizing and leadership outside of the “official” horizontal structure. Morozov links this to a general Internet-inspired delusion in the utility of decentralization, but the problem goes back far before the Internet. Andrew Cornell describes similar observations in his book Oppose and Propose, a history of Movement for a New Society, which was a key progenitor of contemporary horizontal organizing practice in the late 1970’s and 1980’s. One of Cornell’s interviewees from within MNS describes this as “covert leadership”:

At one point, an organizational development consultant volunteered to work with MNS because it seemed as an organization we were getting sick. She had us do an exercise where she said, “All of you who are leaders in the organization, you go over there.” So like three people, all blushing, go across the room. And she smiled and said, “OK, all of you who do covert leadership, you go over here.” And about a third of the room gets up…. So it turned out there was there was this group of covert older male leadership — and this is so traditionally male, too, like we’re holding the family together. – George Lakey, p 73, Oppose and Propose

While Morozov never says it directly, he implies that these failings are a fundamental fault with horizontal models, and that abandoning those models in favor of hierarchy or centralization would be preferable. But I think he misreads Jo Freeman if he believes that the solution to the Tyranny of Structurelessness is to resort to hierarchical organizing. Instead, I would argue that we need a structured, policy-driven horizontalism that works to distribute power equitably, while retaining effective leadership.

Like many before him both inside and outside activist circles, Morozov conflates the notions of hierarchy, centralization, and leadership on the one hand; and horizontalism, decentralization, and leaderlessness on the other. Here’s how I distinguish them:

Hierarchical / horizontal refers to how decision making power is distributed within an organization
Centralization / decentralization refers to how resources and infrastructure are distributed
Leadership / leaderlessness refers to how vision, ideas, and influence are distributed

As Morozov rightly points out, none of the systems that we typically think of as “decentralized” are entirely distributed in their function. Wikipedia may have distributed peer production of content, but it has centralized allocation of servers, and is heavily influenced by the thought leadership of Jimmy Wales and the Wikimedia Foundation’s board of directors. BitTorrent, a paragon of peer-to-peer technology, has decentralized computing infrastructure (no central servers), and has multiple compatible free and open source implementations (no central application); but even BitTorrent is centralized on the level of the protocol: the official specification is controlled by a for-profit company, BitTorrent, Inc. A fun exercise is to pick any ostensibly peer-to-peer or distributed technology, and walk through the OSI Model to see how hierarchies are or aren’t at work at each layer of the technology. Even the Occupy movement, as radically non-hierarchical as it was, centralized on a particular set of meeting techniques for decision making in assemblies, and centralized the resources of activists in single encampments in each major city.

Progressive activists often avoid hierarchical structures because of their role in maintaining systems of inequality and privilege. But as Morozov, Cornell, Freeman, and many others have pointed out, without any structure, we fall back to the same old problematic power imbalances that we began with. What we need is a more nuanced view that recognizes where and why we choose structures of different types to support our goals for equitable and democratic systems.

The Role of Leadership

In Freedom is an Endless Meeting, a history of democracy in American social movements, Francesca Polletta describes the problematic relationship that many groups have had with the notion of “leadership”:

Equality has sometimes been interpreted as prohibiting any differences in skills or talents. What group members have viewed as effective leadership at one point has come later to be seen as manipulation. – Freedom is an Endless Meeting, [p 4]

Is leadership just another form of hierarchy and dominance? Skilled charismatic leaders, such as Martin Luther King, Jr. or Mohandas Ghandi, can be of tremendous value to a movement by providing a unified vision and set of values around which people can organize. However, they are also a liability: they concentrate the power of influence into one fallible person, limit the opportunities for new ideas, and can be a “single point of failure” which can end the organization if the leader is arrested, deposed, or killed. But charismatic leadership is not the only way.

Ella Baker, one of the founders of the Student Nonviolent Coordinating Committee (a highly influential civil rights organization founded in 1960) argued against the leadership model expressed by the “prevailing messianic style of the period”, in favor of a model of leadership that focused on developing capacities and skills in others. Baker described this as “group-centered leadership.” Instead of expressing leadership through a strong public personality with a single dominating vision, a group-centric leader helps to develop the capacities of other group members as leaders. Group-centric leaders work with groups to develop group values and group visions, by elevating each member of the group. Later groups formalized this approach into a model for consensus-oriented decision making, and formed what has been described as “leaderless” organizations — organizations that have no single elected representative or charismatic figurehead.

However, this use of the term “leaderless” can lead one to conflate decision-making power (functional leadership, or hierarchy) with the power to share one’s visions, ideas, and values (thought leadership). That mistake leads us down the path to covert leadership or group ineffectiveness. If we think of leadership as a zero-sum commodity, where one person’s expression of leadership limits others’, we can fall into the trap of stymying any efforts by group members to express their skill and vision. Under a group-centric model, leadership is not a zero-sum game: more leadership results in a more powerful and more capable organization, both on the individual and group level.

Can Ella Baker’s group-centric leadership model rescue us from the pitfalls of both horizontalism and hierarchy? I won’t be so “solutionist” as to claim that I know definitively. The models of horizontal decision making in wide use in activist movements today have roots that go back to Quaker traditions, and have been under ongoing refinement as tools for organizing since the civil rights era. Ironically, what I’m arguing here is a more conservative approach than what Morozov suggests: rather than doing away with 50 years of horizontal organizational development due to the supposed failings of one movement, I believe that incremental improvements to the model can help us to correct those failures, without sacrificing what we’ve gained. We can avoid both a naive fetishization and a reactionary aversion to decentralization by recognizing that it can be beneficially used in some domains, and harmfully in others. These things are not set in stone, and as long as we can keep our eyes on the ball of our goals for liberation, we can adapt and improve our tactics.

What currently excites me are organizations that have structures and policies which ensure that power is distributed equitably, but which still celebrate the skillful contributions of leaders. A few examples of some of the ways in which this is accomplished:

Rotation of facilitation roles, giving each member of the group practice in exercising leadership.
Providing free childcare at meetings and conferences, to ensure people with families can attend.
Sliding-scale policies for rent, dues, and travel reimbursements to ensure that people are not excluded for lack of funds.
An explicit focus on education and leadership development as a core part of the group’s mission and operations. This includes regular trainings for group members, as well as teaching others outside the groups.
Celebration of positive examples of leadership through award ceremonies or other forms of recognition.

These are all steps in the right direction. We can look to organizations like the North American Students of Cooperation, the AORTA Collective, the New Organizing Institute, and the Allied Media Project as examples of organizations that are seeking to take this to the next level. [Disclaimer: I serve on the board of NASCO]

Like Francesca Polletta, Andrew Cornell, Jo Freeman, and other thinkers who have taken hard looks at horizontal organizing, I believe that it’s important for groups to use well-structured processes to avoid the pitfalls of the past. To me, this means written, explicit governance policies; though others might find other expressions. Occupy used a lot of successful structures from the past (facilitation structures, hand signals, progressive stacks, well-defined thresholds and definitions of blocks), but also repeated some mistakes (aversion to leadership, ill-defined group membership, lack of a baseline of shared values).

A careful analysis of where and how we seek to distribute power within our groups can help to avoid the pitfalls of uncritical decentralization. Occupy found tremendous value in centralizing on the tactic of encampment, centralizing their physical resources within the encampments, and centralizing deliberation formats through General Assemblies and working groups. But at the same time, they rightly decentralized the power to make decisions. Throwing out decentralization won’t help us; we need to make critical choices about where and when to centralize and where and when to distribute to ensure effectiveness. We can keep the best parts of a decentralized and horizontal organization, and still have good leadership.

Thanks to Ethan Zuckerman, Catherine D’Ignazio, Rodrigo Davies, and others from the Center for Civic Media for giving very helpful feedback on drafts of this post.