Recent policy discourse suggests that arts and humanities research is seen as being less useful to society than other disciplines, notably in science, technology, engineering and mathematics. The paper explores how this assumption's construction has been built and whether it is based upon an unfair prejudice: we argue for a prima facie case to answer in assuming that arts and humanities research's lower societal value. We identify a set of claims circulating in policy circles regarding science, technology, engineering and mathematics research and arts and humanities research's differences. We find two groups: arts and humanities research is less useful than science, technology, engineering and mathematics, and arts and humanities research is merely differently useful. We argue that empirical analysis is necessary to disentangle which ones are true to assess whether policy-making is being based on rational and evidence-based claims. We argue that debates about public research value should recognise that humanities have different (but equally valid) kinds of societal value.