Short abstract

Targeted research funding aims to address societal challenges or needs. We study how targeted funding programmes shift the research landscape. Based on data from SNSF, RCN and NNF combined with Dimensions, we use a high-resolution topic clustering model to compare pre- and post-award publications.

Long abstract

We compare targeted and non-targeted funding programmes to try to understand how targeted research funding shifts the research landscape.

Our analysis is based on data collected from the Swiss National Science Foundation (SNSF), Research Council Norway (RCN) and Novo Nordisk Foundation (NNF), combined with grants and publications data from Dimensions. The collected data spans 100 different targeted funding programmes, covering over 10,000 grants, and spans 70 different non-targeted funding programmes, covering over 80,000 grants. For each grant, we collected publications from the PI, from both before the grant and after the grant. Each publication is assigned to a scientific topic on the basis of a clustering method of the entire citation network of Dimensions.

We study the difference between the propensity to publish in a certain topic before a grant was awarded and the propensity to publish in that topic after the grant was awarded. Due to sparsity, we regularise the calculation of these differences using weakly informative Bayesian priors. The scientific landscape is dynamic and shows endogenous changes, also in the absence of targeted research funding. We therefore compare these differences to baseline rates of change for non-targeted research funding. We aggregate these differences across topics and funding programmes to analyse whether targeted funding programmes show larger changes in the research landscape than non-targeted funding programmes.

We have now collected most data. Initial results suggest there is a close alignment between publishing practices and targeted research funding. We expect to present full results at the conference.