On November 9, 2026, researcher Christian Leibel released an arXiv preprint demonstrating that rising publication counts among early-career researchers reflect expanding co-authorship teams rather than higher individual productivity. Analysing publication records from over 200,000 scientists across five major disciplines between 1996 and 2024, the study shows that once credit is divided among co-authors, individual output remains flat or declines, challenging hiring committees and research evaluators to abandon unadjusted raw publication tallies.
What happened
In academic hiring and grant allocation, a pervasive assumption holds that early-career researchers must publish at substantially higher volumes than earlier generations to remain competitive. Yet whether individual researchers are producing more scientific work or simply sharing credit across larger teams has rarely been tested with longitudinal data. Sociologist Christian Leibel of the Max Planck Society and LMU Munich addressed this question by examining bibliometric metadata from the German Kompetenznetzwerk Bibliometrie, triangulating disambiguated author profiles across OpenAlex and Scopus. The study compared full-count publications—where each paper yields one credit—against fractional counting, which divides credit among co-authors.
Methodology and key findings
The study followed cohorts beginning their careers between 1996 and 2015 over their first ten publishing years across five disciplines: chemistry and materials science, earth and environmental sciences, engineering, health sciences, and life sciences. The core analysis covered 90,735 author records, with multiverse specifications testing up to 211,859 researchers across sixteen model variations of disambiguation thresholds and credit formulas.
The findings demonstrate a clear divergence between raw paper tallies and collaboration-adjusted output:
- Raw paper counts grew: Comparing the 1996–1999 cohorts to the 2011–2014 cohorts, average publication lists rose by 22% in life sciences, 39% in earth and environmental sciences, 40% in health sciences, 44% in engineering, and 46% in chemistry and materials science (one to two additional papers per researcher).
- Co-authorship expanded faster: Over the same period, average team sizes per paper increased by 39% in life sciences, 46% in earth and environmental sciences, 47% in engineering, 56% in health sciences, and 62% in chemistry and materials science, adding two to three contributors per paper.
- Fractional output stayed flat or fell: Adjusting for team size eliminated the apparent productivity surge. Fractional output was flat in health sciences (+1%), earth and environmental sciences (-4%), and engineering (-3%). In life sciences, team growth caused a statistically significant 9% decline in fractional output. Only chemistry and materials science recorded a modest 8% rise.
Why it matters for research practice
These findings show that unadjusted bibliometrics misrepresent individual scholarly productivity. Treating each co-author as a full producer accelerates h-index inflation and rewards researchers who participate in large networks over those conducting focused investigations in smaller teams.
The data also challenges common assumptions about research integrity. Observers often view lengthening junior CVs as evidence of salami-slicing, or dividing projects into minimal publishable units. However, if researchers were systematically fragmenting unchanged workloads, individual fractional output would still increase. Because fractional output has remained flat or fallen, expanding author lists reflect greater project complexity and specialization rather than artificial publication multiplication.
What changes for researchers and institutions
For search committees and funding agencies, the preprint reinforces the need to implement DORA and CoARA research assessment commitments. Evaluating researchers by unweighted paper tallies penalizes scholars in small-team fields while rewarding passive co-authorship. Panels must replace raw volume metrics with qualitative evaluation of substantive contributions.
For early-career researchers, the findings highlight the value of explicit role documentation. When papers involve numerous contributors, publishing a transparent CRediT author statement ensures committees can evaluate specific operational contributions—such as data analysis, methodology, or software—rather than relying on diluted co-authorship status.
