The prevalence, impact, and nature of big team science

Big team science endeavors represent the largest and most ambitious investments of social capital in science. Yet, we lack a systematic understanding of their prevalence, impact, and nature. Here, this gap is filled via an analysis of >29 million papers published in 19 disciplines over 200+ years. Results indicate that big team science is increasingly frequent – yet extremely rare – and generally focuses on building upon (vs. disrupting) lines of research. Consistent with collective intelligence theories, unusually large teams tend to have unusually large impact, in terms of median yearly citations. However, such teams navigate a diseconomy of scale, wherein adding co-authors yields diminishing increases in citations. An analysis of contributorship classifications from > 47,000 papers further illustrates how the structure of scientific collaborations evolves in pursuit of such efforts. Consistent with structural differentiation theories, increasingly large collaborative efforts feature increasingly specialized roles. Taken together, results provide a systematic, evidence-based characterization of big team science: unusual endeavors wherein researchers typically adopt specialized roles and navigate inefficiencies to make unusually high impact advancements to existing lines of thinking.