The Innovation Shock: What Happens When a Key Collaborator Disappears?

Innovation is often celebrated as the triumph of brilliant individuals. In reality, modern research and development is deeply collaborative: patents are rarely filed alone, breakthroughs increasingly emerge from teams, and knowledge flows across organizational boundaries. When one link in that chain disappears, what happens to the rest?

A new study published by Felix Poege (Department of Management and Technology and ICRIOS, Bocconi University), Fabian Gaessler (Universitat Pompeu Fabra, Barcelona), Karin Hoisl (University of Mannheim, Germany), Dietmar Harhoff (Max Planck Institute for Innovation and Competition, Munich, Germany) and Matthias Dorner (Institute for Employment Research, Nuremberg, Germany) in Management Science offers one of the most rigorous answers to date. It shows that losing a collaborator inside the firm can often be absorbed. Losing one outside the firm is another story. At the center of the research is a stark empirical strategy: the unexpected death of inventors.

From organizational context to a natural experiment

The study begins from a simple but overlooked observation: most knowledge workers are embedded in organizations. Yet prior research on collaborator loss has paid limited attention to whether a lost collaborator was internal or external to the firm. The authors argue that the effect of collaborator loss on individual productivity depends on whether the lost collaborator is internal or external to the remaining knowledge worker’s organization.

To study this distinction, the researchers rely on the INV-BIO dataset, a unique panel covering more than 150,000 German inventors between 1980 and 2014. This database combines administrative labor market records tracking exact employment relationships with detailed patent information from the European Patent Office and the German Patent Office. It allows the authors to identify not only who collaborated with whom, but also whether collaborators were working in the same establishment at the time of death.

The empirical challenge is formidable. Productivity could decline for many reasons unrelated to a collaborator’s loss: firm strategy shifts, life-cycle patterns, technological shocks. To isolate causality, the authors exploit 845 unexpected deaths of active inventors aged 60 or younger. These deaths are treated as plausibly exogenous shocks.

Each deceased inventor is carefully matched to a “pseudo-deceased” inventor with similar observable characteristics (age, gender, technological field, lifetime patent count, and employer size). The remaining collaborators of the deceased ones form the treatment group; collaborators of the matched pseudo-deceased form the control group. A difference-in-differences framework then compares productivity before and after deaths across these groups. This design ensures that what is being measured is not a gradual decline in a team, but the discrete impact of losing a specific individual.

The asymmetry that changes the story

At first glance, the overall result aligns with prior literature: collaborator loss reduces inventive productivity. Over eight years, patent output falls modestly.

But the aggregate masks a sharp asymmetry: when the lost collaborator was external to the firm, productivity drops significantly—about 8 percent in simple patent counts and 14 percent when weighted by citations. When the collaborator was internal, the effect is practically zero.

This is not because internal collaborators are less important. The authors show that even when focusing on collaborators with highly complementary knowledge, large networks, or intense collaboration histories, the pattern persists. External loss hurts; internal loss is muted.

Firms can fill some gaps but not others

The explanation lies in the organization’s ability to respond. Firms invest in knowledge management systems that codify and disseminate expertise internally. They also possess hiring capabilities and internal labor markets that allow them to reconfigure teams when needed. These mechanisms can “fill the gap” created by an internal collaborator’s death.

Indeed, the study shows that in firms with strong knowledge management and hiring capabilities, internal collaborator loss can even be followed by higher productivity. The authors read this as evidence of successful reallocation or upgrading of human capital. However, this resilience has limits. “The loss of a high-performing internal collaborator has a significant negative effect”, they admit, suggesting that exceptional talent cannot always be replaced.

External collaborators present a different challenge. Firms do not control the knowledge management systems or hiring processes of other organizations. They cannot easily access the replacement collaborator who fills the gap elsewhere. As the authors note, “knowledge workers can hardly draw on the knowledge management of other organizations or immediately build a tie to the individual filling the gap created by the lost collaborator.”

Implications for innovation strategy

This study suggests that peer effects may appear weak within firms not because colleagues do not matter, but because organizations actively cushion the blow of internal shocks. It also highlights a structural vulnerability in open and cross-boundary innovation: the more firms rely on external ties, the harder it becomes to hedge against their sudden disappearance.

Managers should then not retreat from collaboration, but they should keep in mind that resilience is uneven. Internal networks can be rebuilt. External bridges, once broken, are far harder to reconstruct.