In answer to a question about moving between employment in the UK and the US, user StrongBad suggests that "the obsession with the UK REF will likely push your CV in directions that are not ideal for getting a job in the US."
This point struck me as interesting, as in some respects it appears counter intuitive. When I asked why this was the case, he suggested that it would make a good question. I'm now asking that question.
For those unaware of the REF, it is a regularly occurring (and not uncontroversial) nationwide assessment exercise which is used to guide the distribution of funding amongst UK universities.
In order to perform well in the REF, there is a pressure on researchers in the UK to produce a number of publications (four, to be precise) which are deemed to be of a high level. In assessing quality, reviewers are guided to consider whether a piece of work is "world-leading in terms of originality, significance and rigour". Furthermore, since the most recent REF another consideration for assessment has been whether research has "an effect on, change or benefit to the economy, society, culture, public policy or services, health, the environment or quality of life, beyond academia."
What's therefore not clear to me is why would building a CV containing publications deemed by peers to be world-leading, and conducting research which peers deem to have benefit outside of academia, not be ideal for job prospects in US academia?
I believe you are correct, and that the statement: "the obsession with the UK REF will likely push your CV in directions that are not ideal for getting a job in the US." is mostly incorrect.
On the contrary, the UK REF is likely to have a good impact on the publication habits of UK scholars, as it specifically favors small amount of very high quality (i.e., 4 per four/five years) publications, over high amount of medium or low quality publications.
For instance, to my understanding, for the REF and excellent paper is possibly a technical paper. E.g., a paper that merits publication in the Annals of Mathematics, or Trans. of the AMS in the case of a math department. To my knowledge it is not meant to be papers that are, say, Nature equivalents. Indeed, you can see that many schools according to the 2014 REF metric had produced many top papers (and had it been required to have Nature equivalent publications this would not have been possible).
Note that in the US you will find it hard to get into a good research school without at least some such top papers.
One thing that may hinder a bit basic research is the need to have "impact cases", i.e., research that leads to practical/social/policy-wise impact. But not every academic should have an impact case, and so this may not be a very big problem.