Knowledge-integrated research is more disruptive when supported by homogeneous funding sources: a case of US federally funded research in biomedical and life sciences

Governments commonly support scientific research integrating knowledge of various areas and fields, expecting this to lead to disruptive research. However, empirical evidence is relatively scant regarding whether knowledge-integrated research with government funding disrupts sciences. This study explores the quality of government-funded research in terms of knowledge integration and disruption. Knowledge integration is measured as the degree to which a paper combines different research subjects in three dimensions of diversity—variety, balance, and disparity. Disruption is calculated based on paper citation networks, thereby capturing how the paper challenges existing research streams and provides new directions of scientific research. Our analysis demonstrates that the relationship between knowledge integration and disruption varies by government funding sources, using bibliometric data from US federally funded research published between 1975 and 2005 in biomedical and life sciences. The empirical findings show that: (1) knowledge-integrated research is more disruptive when supported by funding agencies of the same government department rather than of different government departments; and (2) while the effect of knowledge integration on long-term citations is positive and becomes stronger for the studies funded by homogeneous government funding sources, its effect on short-term citations is negative regardless of the source of government funding. These results highlight the potential value of knowledge-integrated research that is hard to capture through short-term performance, which holds significant policy implications for current practices in research evaluations that are driven by short-term evaluation.