Team Size and Collaboration

The previous chapter, Gender and Patenting, examined how gender composition in patenting has evolved. This chapter turns to the structural organization of inventive activity itself: the shift from solo to team-based patenting, the distribution of solo invention across technology fields, and the relationship between team size and patent quality.

The most consequential structural shift in modern invention has been the transition from individual to collaborative patenting, reflecting the growing complexity and interdisciplinarity of contemporary technology development. This chapter traces that shift and explores whether team composition correlates with measurable differences in patent quality.

Average team size increased from 1.7 inventors per patent in 1976 to 3.2 by 2025 (Through September). The share of solo-inventor patents has declined substantially — from 58% to 23% — even though absolute solo patent counts nearly doubled (from ~41,000 to ~77,000 in 2024), as team-based patents grew far more rapidly. Meanwhile, patents listing five or more inventors have become increasingly prevalent, growing more than eightfold from 2.5% to 21%.

Solo invention rates vary substantially across technology fields. Chemistry and metallurgy (CPC Section C) has the lowest solo-inventor share at 13.2%, reflecting the laboratory-intensive, multidisciplinary nature of chemical and pharmaceutical innovation. Fixed constructions (Section E) retains the highest solo share at 39.0%, consistent with fields where individual craftsmen and small-firm inventors remain more prevalent. Physics (Section G) and Electricity (Section H), the two largest and fastest-growing technology domains, both exhibit solo rates below 26%.

Team size categories are defined as: Solo (1 inventor), Small (2–3 inventors), Medium (4–6 inventors), and Large (7+ inventors). The following quality indicators reveal how team composition correlates with patent characteristics.

The descriptive quality comparisons above reveal suggestive patterns but cannot isolate the independent effect of team size from confounding factors such as technology field, grant year, and assignee characteristics. To address the limitation, the following analysis applies a Frisch-Waugh-Lovell (FWL) OLS regression that demeans cohort-normalized 5-year forward citations within grant year, CPC section, and assignee-size groups. Team size dummies (2–3, 4–6, 7+ inventors) are estimated relative to the solo-inventor baseline, with standard errors clustered at the assignee level.

Beyond team size, the institutional diversity of inventor teams provides an additional dimension of collaboration. Cross-institutional patents — those listing inventors from more than one assignee type (such as a corporation and a university) — represent a distinctive form of collaboration that bridges organizational boundaries.

The team-based structure of modern invention documented in this chapter concludes ACT 3: The Inventors. Having examined who invents — from top inventors and specialization patterns to career trajectories, gender composition, and team structure — the analysis now shifts to where innovation happens. ACT 4: The Geography begins with Domestic Geography, which maps the concentration of patent activity across US states and cities, revealing the spatial clustering that shapes patenting activity in the United States.