‘Existential risk’ to research from failure to demonstrate impact

The issues here are the same as they have always been. 1. If impact is viewed as something that happens directly, then 'upstream' research (e.g. pure maths) will be pushed aside because it is so difficult and so long-term to track impact a long way downstream. This would be a mistake even assuming a narrowly pragmatic view of impact; even more so if we allow slow-burning impacts on other aspects of a whole culture. 2. Impact may not be 'popular'. The same arguments as might be deployed against, for example, arts museums, can be deployed against much research. Is this a good idea?

We know that we've made a mistake by introducing all sorts of short-term indicators of research productivity, such as citations, number of publications, the h-index. But the suggestion to replace these with another short term measure, that is direct impact, is just failure to learn from our mistakes. I also note that promotion procedures reward fame and influence in the academic community, which is yet another mistake, as it rewards toxic and narcissistic behaviour. I mean some promotion criteria are just straight out of the DSM.