Impact of bibliometrics upon the science system: Inadvertent consequences?

The introduction of bibliometric (and other) ranking is an answer to legitimation pressures on the higher education and research system. After years of hesitation by scientists, science administrators and even politicians in many of the industrialized countries, the implementation of bibliometrics based (and other types of) rankings for institutions of higher education and research is now being introduced on a full scale. What used to be an irritation to the parties concerned has suddenly become a fad. In contrast to this rather sudden enthusiasm, there is very little reflection on the impacts of this practice on the system itself. So far empirical data on the impact of bibliometric rankings seem to be available only for two cases: Australia and the British research assessment exercise (RA-E). Thus, the actual steering effects of bibliometric rankings, the reactions of the system are largely unknown. Rankings are in urgent demand by politics. The intended effect is to create competition among institutions of higher learning and research and thereby to increase their efficiency. The rankings are supposed to identify excellence in these institutions and among researchers. Unintended effects may be 'oversteering', either by forcing less competitive institutions to be closed down or by creating oligopolies whose once achieved position of supremacy cannot be challenged anymore by competitors. On the individual level the emergence of a kind of 'chart' of highly cited stars in science can already be observed (ISI HighlyCited.com). With the spread of rankings the business administration paradigm and culture is diffused through the academic system. The commercialization of ranking is most pronounced in the dependence of the entire practice on commercial providers of the pertinent data. As products like IST's Essential Science Indicators become available, their use in the context of evaluation tasks is increasing rapidly. The future of the higher education and research system rests on two pillars: traditional peer review and ranking. The goal must be to have a system of informed peer review which combines the two. However, the politicized use of numbers (citations, impact factors, funding etc.) appears unavoidable.