A structural model of coronavirus behaviour for testing on data behaviour

ABSTRACT

We fit the logistic function, the reduced form of epidemic behaviour, to the data for deaths from Covid-19, for a wide variety of countries, with a view to estimating a causal model of the Covid virus’ progression. We then set up a structural model of the Covid virus behaviour based on evolutionary biology and social household behaviour; we estimated and tested this by indirect inference, matching its simulated logistic behaviour to that found in the data. In our model, the virus’ progression depends on the interaction of strategies by household agents, the government and the virus itself as programmed by evolution. Within these interactions, it turns out that there is substitution between government topdown direction (such as lockdown) and social reaction to available information on the virus’ behaviour. We also looked at the experience of second waves, where we found that countries successfully limited second waves when they had had longer first waves and followed policies of localized reaction in the second.

KEYWORDS:

• Coronavirus
• Covid-19
• evolution
• optimization
• indirect inference
• lockdown

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

¹ Source of data: Johns Hopkins University Center for Systems Science and Engineering

² https://www.biorxiv.org/content/10.1101/2020.05.21.108308v1.full

³ https://www.cell.com/action/showPdf?pii=S0092-8674%2820%2930610-3

⁴ The J-value (cost per life saved by safety measures) used by the US Dept of Transportation is between £ 4 and £ 10 million per life saved through road safety measures, with the UK value being around £ 9 million.https://www.bristol.ac.uk/media-library/sites/policybristol/PolicyBristol-Report-April-2018-value-human-life.pdf