Excellence and growth dynamics: A comparative study of the Matthew effect

In the past two decades, centres of excellence (CoE) and other 'research excellence initiatives' likely to increase the cumulative advantages and stratification of science, have been implemented in many countries. Based on empirical studies of CoE in four Nordic countries, this paper examines how the resources provided by CoE schemes (generous long-term funding, prestige and visibility) add to the success and growth dynamics of the CoE. The data indicate a modified Matthew effect with ceilings and limits avoiding excessive accumulation of resources. Important impacts of the CoE are found, in particular in terms of enabling more interdisciplinary collaboration and risk-taking and enhancing international recruitment to the research areas involved. But, in contrast to what might be expected, the CoE grant seem to add less to the relative citation rate of those already performing at the highest level, than for those performing at a somewhat lower level prior to the CoE grant.