A different type of modelling could have been used during Covid - Yorkshire Post Letters

The Covid inquiry’s second report from Baroness Hallett and her team gained several columns in November’s newspapers. The report’s findings apparently include concern over the Government’s delayed start date to the first lockdown of March 2020, a failure to acknowledge the need for urgency in managing the Covid virus, and groupthink among the advisory agency SAGE – which recommended an even later start date for the government.

Unsurprisingly in the field of healthcare, mathematical models were used to plot and predict, each model aiming to reflect in numbers some aspect of how the world is, or how it might soon become. Similar models have been applied in the commercial world.

However, there is a radically different type of model which could have been of service during the problematic Covid years, a type developed over a 30-year period by researchers based at Lancaster University’s Department of Systems Engineering.

These models did not claim to be scientific, just as, say, most colleges do not claim to follow scientific practice when teaching history, art or business studies.

Professor Peter Checkland joined the Department of Systems during the late 60s to support its consultancy role in the real-world outside academia. His brief aimed at improved understanding and action when clients admitted they were burdened with obscure problems: yes, they had issues to resolve, but they could not agree in detail on just what those issues were.

What emerged from Checkland’s participative work on behalf of hundreds of clients was a set of principles named Soft Systems Methodology (SSM) to guide coherent inquiry within difficult settings. With only four stages, the recursive methodology is easy to learn.

First, gain an appreciation of the setting which is under review; second, draft systems models which seem relevant to alleviating unique concerns; third, use the models as a base for orderly questioning of organisational staff (at best, all will be invited to join the inquiry); and then, in stage four, changes agreed from stage three can be actioned.

To briefly introduce the thinking required for this approach, the accompanying SSM model shows a purposeful set of phrases that propose - in logic - protecting social groups as the way to counter a virus. Not an exercise in design, the model aims to sharpen discussion when ‘the problem’ cannot be defined clearly – a changeful mix of opinion, doubts, and politics. Each of the phrases (‘activities’ in SSM’s jargon) shown is subject to questions which have proved practical when sharing the methodology:

How is this activity (Assess the nature of virus, etc.) carried out? If not carried out, ought it to be? How is performance measured? Who is in charge? What would count as improvement?

Although activity four, define affordable, perhaps reads as philosophical, a major response by health institutions will make demands on their limited person-hours, vehicles, and other resources.

Like the model per se, each activity persists through time; exactly how activities get done – phone calls, video conferencing, delegating, inviting expert opinion from Europe - depends on convention and the personnel involved.

SSM does not start an inquiry with traditional so-called information systems (a misnomer) but instead shows images which are taken to be notional missions. And in contrast with the mysterious unseen workings of AI, the logic here is explicit, open to criticism, and easily replaced by more relevant ideas at any time: ideas displayed as SSM models.

Any model failing to stir serious parley should be binned; ironically, the word system should also have been dropped from this methodology decades ago since it causes confusion by implying devices and organisations which exist in the social world – while Checkland’s systems try to help explore that multifaceted, unfathomable world.