Short abstract

Why do researchers peer review? Analyzing administrative data from physics journals, we find that reviewers learn by reviewing: They are over twice as likely to cite reviewed papers.

Long abstract

Why do researchers contribute their time to peer review despite increasing pressures on their schedules? While previous research has emphasized altruistic motives and reputational benefits, we propose that personal learning represents a significant yet overlooked incentive. We investigate whether and when researchers acquire knowledge by peer-reviewing using administrative data from the Institute of Physics comprising 104,306 reviewer-manuscript pairs from 60 peer-reviewed journals.

Our analysis, involving within-manuscript comparisons between reviewers and those who were invited to review but were unavailable, reveals that reviewing a manuscript more than doubles the likelihood of citing it within three years (1.3% → 3.1%, p<0.001). Notably, this effect is moderated by both geographic and intellectual proximity. While invited reviewers from the same country as the manuscript's corresponding author who do not review cite the paper at high rates (3.0%), invited reviewers from a different country who do review cite the paper similarly, suggesting that reviewing compensates for geographical barriers. Reviewing has its largest effect on citing when the paper is intellectually very close to the reviewer’s recent work, suggesting that peer reviewing does little to break intellectual barriers.

Overall, reviewers learn by reviewing, helping explain their continued participation in the process. The results also illuminate how knowledge diffuses across geographic and intellectual boundaries.