The ability of universities to efficiently produce high-standard knowledge has become an important goal in science policies of many developed countries. Thus, many countries nowadays steer universities based on performance monitoring and competition. This article analyzes the connection between the competitiveness of the university funding environment and research performance in five OECD countries in 1987-2006. Besides funding, other science policy factors are analyzed using the framework of four state steering models. Results indicate that the university funding environment has become more competitive in all the compared countries, but the extent and pace of this development varies. Countries also differ in relation to steering models but all have employed policy elements typical of at least two models. In terms of competitive funding environment and research performance, there is no straightforward relationship between the two. With reference to the state steering models, the most traditional model which emphasizes university independence from the state seems to be the most beneficial to research performance. NOWLEDGE HAS BECOME an important resource for industry as an element of both products and production processes over the last couple of decades. This development is associated with the globalization of economy. The developed, industrialized countries have faced a challenge of maintaining their position in global economic competition. Universities have become key components of the economies of developed countries because they create knowledge and disseminate it to industry and wider society and provide the highest education to people. Because of the strategic significance of universities, their performance and costeffectiveness have become highly relevant issues to policy-makers (Slaughter and Leslie, 1997: 36-40).As a result, there is a strong science policy trend that emphasizes the research performance of the university sector. Competition for money and other financial incentives are often used as steering instruments, since funding is regard as having a strong impact on the behaviour of universities which are dependent on resources from other organizations such as state agencies (Hackett, 1990; Nieminen, 2005: 124-125). However, this development is not uniform across countries. Also the long-term usefulness for and impacts of funding incentives and competition on university research are still unclear.In this article, we analyze the connection between the funding and research performance of universities. In addition, we scrutinize the role of other K Laura Himanen, Otto Auranen (corresponding author) and Hanna-Mari Puuska are at

Monitoring of research performance, especially performance-based allocation of research funding inevitably creates intended and unintended incentives for universities. In this article, we study if monitoring performance acts as an incentive for improved research performance by scrutinizing the development of two essential indicators of Finnish universities’ research performance, publication output, and international competitive research funding at both national and organizational level, by combining several statistical data sources. The results are reflected to the recent changes in Finnish universities’ national funding model as well as the organizational research assessment performed in 12 universities between the years 2009 and 2019. Based on our analysis, we suggest that the incentives brought on by national and organizational level actions have not hindered the positive development in the two elements of research performance as seen in the data, and in fact the introduction of an indicator that emphasizes the quality of publication channels to the university funding model has incentivized positive development in publication counts. Universities’ research assessments seem to have no systematic influence on the developments of the two indicators under scrutiny. However, according to our analysis, research assessments in Finnish universities are conducted first and foremost for developmental reasons, so it is likely that possible changes in research performance happen during a much longer period than what our data cover.