Bibliometric evaluation vs. informed peer review: Evidence from Italy

A relevant question for the organization of large-scale research assessments is whether bibliometric evaluation and informed peer review yield similar results. In this paper, we draw on the experience of the panel that evaluated Italian research in Economics, Management and Statistics during the national assessment exercise (VQR) relative to the period 2004-2010. We exploit the unique opportunity of studying a sample of 590 journal articles randomly drawn from a population of 5681 journal articles (out of nearly 12,000 journal and non-journal publications), which the panel evaluated both by bibliometric analysis and by informed peer review. In the total sample we find fair to good agreement between informed peer review and bibliometric analysis and absence of statistical bias between the two. We then discuss the nature, implications, and limitations of this correlation.