Abstract
Firms frequently publish scientific articles as part of their R&D initiatives, motivated by commercial objectives. However, the extent of industry involvement in publishing varies across different scientific fields and can have implications for research within those fields. Novelty in science is associated with scientific and technological breakthroughs. In this paper, we examine a field-level antecedent of novelty-the extent of industry publishing contribution to a field-and its association with two key aspects of recombinant novelty of publications: the occurrence of a novel recombination (novelty occurrence), and the degree of novelty, captured through novelty breadth, reflecting the scope of novel integration of knowledge elements, and novelty distance, reflecting the extremity of conceptual divergence among novel knowledge recombinations. Drawing on a longitudinal dataset of 11.1 million publications across 1639 STEM fields from 2000 to 2014, we find that greater industry publishing contribution within a scientific field is associated with higher odds of novelty occurrence and greater novelty breadth, but lower novelty distance. Notably, university is an important driver of our results across the three dimensions of novelty highlighting the importance of industry publishing contribution in shaping the novelty of entire fields. In addition, we find that top-ranked research institutions appear better able to manage the trade-off between novelty distance and other forms of novelty as industry publishing contribution increases. Our findings emphasize the need for policymakers in encouraging and preserving more exploratory forms of novelty in fields with substantial industry publishing, where such exploration is particularly valuable.