The great, late, regulation debate

In their surprisingly readable A short guide to regulation the National Audit Office sets out a scale of all the ways in which regulation can work.

At one end of the scale is command and control regulation. This includes mandatory government guidance and legal penalties. At the other end of the scale there is no-intervention at all. This means letting the market decide how things turn out. In between, there are codes of practice, information campaigns, economic controls, and other regulatory measures.

It’s not that any one measure is better than another, it is that some are more appropriate in some situations than others. For example, it would be less than ideal to have an entirely free market nuclear sector given that an extinction event from lax work is a pretty enormous negative externality.

Regulation today

In research it’s not clear how regulation should work as a system.

There are some things that are illegal. Researching weapons for sanctioned countries. Sharing IP which could be used by a hostile nation. Tricking research participants into situations which could harm them.

There are things that researchers are encouraged to do. There are funding calls, challenge projects, promotion processes, publication metrics, REF, and all sorts of other things that incentivise researchers to do things.

There are also information campaigns to encourage universities to undertake socially useful activities. KEF, for example, is nothing if not a motivation machine for universities to understand how their knowledge-exchange activities match-up to other institutions like theirs.

These are all regulatory measures but this isn’t the same as a regulatory approach to research.

OfS meets REF

The Office for Students is an obvious contrast here. The OfS sets out a regulatory framework covering everything from academic misconduct, to teaching, to financial probity, and then sets out a set of expectations of how universities will meet their regulation.

It is a separate debate whether the regulation is right but the intention is clear. It is a set of things that anybody can point toward and understand what is expected of participants in the higher education sector.

In research it’s just different. It’s more abstract. It’s a bit of tacking between lots of incentives, imperatives, and issues. It’s regulation through practice, or good conduct, or through peer expectation. It is regulation by vibes.

Different paths

In some places there is emerging consensus on how regulation should work. Most obviously, the Nurse review, Tickell review, and ARIA, all look toward models of less regulation in order to allow researchers to get on with research. This consensus speaks to another implicitly accepted truth that research regulation benefits from peer-review and mutually enforced self-regulation. It is more QAA than OfS.

In his independent review of research bureaucracy Adam Tickell calls for the harmonisation of processes, data and systems, Paul Nurse in his eponymous review calls for a more ARIA like system which would inevitably mean dispensing with huge swathes of processes and systems. ARIA itself is a rejection of almost all regulation and prizes researcher time and independence. Simultaneously, UKRI’s Safer and Better Funding Programme looks to standardise and simplify processes across the funding lifecycle.

These approaches are not wholly compatible and the debate shouldn’t be solely whether we want researchers doing more research. The debate should be at what cost we’re willing to allow that version of research regulation to exist. And it is also necessary to be clear that time and funding aren’t the only regulatory mechanisms available.

For example, should the government regulate through its agencies to withhold funding where universities don’t meet socially beneficial outcomes. QR could have an allocation based on the gender gap between the research teams. APPs could cover PGT to PGR progression for underrepresented groups. There could be multipliers of funding for how much research is retained within local or national economies.

In truth, the mechanism is less important at this stage. It is as Nurse said about the policy ecosystem more generally there should be a coherent set of desired outcomes and then regulation should fall behind them.

It’s time to not only decide what should be achieved but how it should be regulated.