Breaking Library Echo Chambers: Why Innovation Depends on Cross-Department Collaboration

Academic libraries often talk about innovation, experimentation, and risk-taking—yet our organizational structures frequently make these aspirations difficult to realize. Clay Shirky’s Here Comes Everybody, particularly the chapters on structural holes, power-law distributions, and what he memorably calls “failure for free,” offers a useful lens for understanding why.

At its core, Shirky’s argument is not about technology. It is about organizational design, transaction costs, and the hidden ways institutions unintentionally suppress good ideas—dynamics that will feel immediately familiar to academic librarians working within complex, risk-averse institutions.

Bridging Beats Brilliance

Drawing on Ronald Burt’s work on social capital, Shirky shows that the highest percentage of good ideas came from people whose contacts were outside their own department. In contrast, people who were highly connected only within their own unit were less likely to produce ideas that translated into organizational value.

As Shirky notes in his discussion of Burt’s findings, dense, inward-facing departmental networks often create an echo chamber effect (p. 231). Ideas circulate, intensify, and become more internally coherent—yet lose relevance beyond the local context. New managers in Burt’s study disproportionately rejected these ideas not because they lacked intelligence, but because they were too embedded in departmental minutiae and offered little strategic advantage to the organization as a whole.

Bridging networks disrupt this pattern. People who connect across departments gain early access to diverse, sometimes contradictory information. That exposure—not individual brilliance—is what produces good ideas. Creativity, in this sense, is not a personal trait but a structural outcome.

The Institutional Bias Against One Good Idea

Shirky also highlights a persistent organizational dilemma: institutions cannot afford people who produce only one good idea. Because of transaction costs—hiring, management, benefits, evaluation—organizations prefer “steady performers” over “brilliant but erratic” contributors.

The result is an 80/20 logic applied to people: organizations optimize for consistent output and quietly discard the long tail of occasional but valuable contributions. Many good ideas never surface, not because they are bad, but because they are too expensive to discover.

Libraries are not immune to this logic. Grant proposals, pilot programs, and new services often require layers of approval, justification, and assessment before they can even begin. By the time an idea is deemed safe enough to try, it has often lost its urgency—or its originality.

Failure as a Feature, Not a Bug

Open systems, Shirky argues, are remarkably tolerant of failure. They do not attempt to prevent failure; instead, they make failure cheap enough to be routine. Open-source communities and peer-production models do not reduce the likelihood of failure—they reduce the cost of failure.

This is what Shirky means by “failure for free.” Most open projects fail. A few succeed modestly. A very small number succeed spectacularly. But because the system does not require centralized approval or predictive filtering, experimentation can happen at scale. Trying something is often cheaper than deciding whether to try it.

Institutions, by contrast, must coordinate labor, resources, and accountability in advance. As a result, they are pushed toward safe choices, even when safety systematically undermines innovation. Open systems outperform not because they choose better ideas—but because they allow more ideas to be tried.

For libraries, this challenges a deeply ingrained mindset. We often attempt to innovate by minimizing risk—through extensive planning, consensus-building, and outcome prediction. Open systems suggest the opposite approach: create conditions where many small experiments are possible, and let value emerge through use.

What This Means for Academic Libraries

For academic librarians, Shirky’s argument points toward a shift from tightly managed innovation to collaboratively coordinated experimentation—a model that aligns closely with values of shared governance, professional trust, and collective expertise.

Libraries are institutions, not open-source communities—but we can still borrow structural lessons. Innovation is more likely when we reduce the cost of trying new things and allow ideas to emerge through use rather than prediction.

Some practical implications include:

• Lowering the cost of pilots. Small-scale trials, time-limited experiments, and opt-in services reduce the fear of failure.

• Designing for contribution, not perfection. Allow staff, students, and faculty to contribute ideas without requiring full institutional buy-in upfront.

• Valuing bridges. Reward cross-unit, cross-disciplinary, and boundary-spanning work—especially when it connects the library to teaching, learning, and student support.

• Accepting uneven participation. Power-law distributions are normal. A small number of contributors will do most of the work, while many others contribute occasionally—and that is not a flaw.

Rather than centrally controlling innovation, libraries can focus on creating conditions where collaboration is possible, coordination is light, and failure is survivable.

• Accepting uneven participation. Power-law distributions are normal. A handful of contributors will do most of the work—and that is not a flaw.

From Risk Management to Learning Organizations

Most institutions try to manage failure by reducing its likelihood. Shirky suggests a different strategy: manage failure by reducing its cost. This shift aligns closely with learning organization principles, where feedback, adaptation, and reflection matter more than prediction.

For academic librarians navigating AI adoption, digital scholarship, student success initiatives, and evolving institutional expectations, the lesson is clear. Innovation does not require better forecasting. It requires better structures—structures that make it safe to try, acceptable to fail, and possible for unexpected value to emerge.

In a world where no one can reliably predict what will work, the most innovative organizations are not the ones that choose best—but the ones that can afford to choose often.