---
title: "Stern and REF: assessing the past to fund the future with non-portability"
layout: post
---
Something that occurred to me about the Stern review of REF and the proposed non-portability of research outputs is how this changes the relationship of funding to researchers vs. funding an environment. I will note that I have [some qualms](https://figshare.com/articles/Excellence_R_Us_University_Research_and_the_Fetishisation_of_Excellence/3413821), in theoretical (but not political) terms, about such future prediction of excellence through peer review. But that's what REF does. 

REF has always been an activity that assesses a _past_ five-year cycle in order to fund a _future_ five-year cycle. However, when researchers can take their outputs with them, what is actually funded is a snapshot moment of activity; one fundable moment. The institution then receives QR for that snapshot for the next five years.

By contrast, if outputs "stick" to an institution, then a slightly different mode of assessment comes into play. A broader sample can be taken from across an entire five-year period by a university and submitted. This seems to help a little with the gaming on the hiring front.

When it comes to the reward of QR, though, I wonder how much it helps? The proposed non-portability of research outputs means that a department could be obliterated and still submitted to REF. If QR was awarded on this basis, it would be out of proportion to continuing to fund the same research. And this seems to be a fundamental (and pragmatic) decision in REF's design: it has to assume that there is parity between the submission and the future need for funding.

I don't have an answer here, but a question for the implementation phase is surely: what can be done to get the positive benefits of eliminating transfer gaming and assessing environment, while also mitigating any risk that could be incurred through university downsizing yet still exploiting non-portability?