The Institute for Fiscal Studies has published research by Banks et al. entitled Recessions and health: the long-term health consequences
of responses to the coronavirus, which investigates the ways existing literature and models on this topic can be applied to the current COVID-19 pandemic, with a specific focus on the UK. In particular, they attempt to apply the model of the impact of economic shocks on morbidity in Britain estimated in Macroeconomic Conditions and Health in Britain: Aggregation, Dynamics and Local Area Heterogeneity by Janke et al. According to the paper, although they note that:
The Janke et al. analysis looks at the prevalence of long-standing
health conditions but does not examine the intensity or the duration
of the condition within an individual’s life course.
They state that
Quantitatively, Janke et al. estimate that a 1% fall in employment
leads to a 2% increase in the prevalence of chronic illness.
Only about half this effect will be immediate: the full effect will
not be felt for two years. The shock to employment from the
coronavirus pandemic is likely to be much larger than this and so we
may expect a larger rise in poor health.
Given that the Office for Budget Responsibility (OBR)'s recent coronavirus reference scenario predicts a rise in unemployment from 4% to 10%, applying this rule would result in a 12% increase in the prevalence of chronic illness.
They also look at the conclusions of Van den Berg et al.'s Economic conditions early in life and individual mortality, which looked at the effect of the state of the economy at birth on one's life expectancy, and concludes that being born in a recession reduces life expectancy by around 5%. They note that this is a result of many indirect effects - for instance:
Currie (2009) documents the extensive evidence on the links between
parental circumstances and child health, as well as the subsequent
link between a child’s early-life health and their eventual
educational and labour market outcomes. It is well documented that
poor nutrition in early childhood, as well as in utero, will have a
long-lasting impact on individuals, and there are many other examples
where vulnerabilities and shocks in early life have long-term
As well as physical health problems as a direct result of the shutdown measures, they also look at the potential indirect effects on mortality as a result of deteriorating mental health. This section again draws on the Janke model:
Estimates drawn from Janke et al. (2020) suggest that if the economic
downturn were similar to that after the 2008 financial crisis, the
number of people of working age suffering from poor mental health
would rise by half a million.
However, they note that using existing models to predict the effect on mental health is particularly tricky in this case:
To add to this, social distancing in itself is likely to have complex
and nuanced effects on individuals’ social isolation and mental health.
Another issue they identify that will make the effect of the economy on mortality more difficult to model is the disproportionate effect of the lockdown conditions on specific industries, for example, the tourism & hospitality sectors. Again using the Janke model, they note that:
Janke et al. (2020) find heterogeneous morbidity responses to economic
shocks across local areas. Those areas that are hit hardest are those
that are the most deprived and have older populations and older
industrial structures, which are precisely the kind of areas least
able to withstand negative shocks.
It is therefore imperative that any accurate model must factor in the specific demographics in the communities which most rely on the specific sectors most at risk from the lockdown conditions.
Finally, on a more positive note, they observe that negative economic shocks have been shown to contribute to a fall in unhealthy habits such as drinking and smoking, and note recent reports, for example in National Geographic, that the lockdown conditions have contributed significantly to lower pollution levels.
Hopefully, this paper shows the level of detail which would have to be attained by any model with any hope of being remotely accurate, and as a result, proves its current infeasibility. As these lockdown conditions are unprecedented, it is impossible to predict with any certainty how, for example, mental health will be affected by an extended lockdown.
Furthermore, unlike the analysis in the model on direct deaths caused by the virus in your question, there is no data on the long-term economical effects of a lockdown specifically on which to test any such model, which contributes to the level of uncertainty.
Despite these difficulties, it does draw some fairly rudimentary conclusions using past research about how the economic effects of the strict measures on indirect mortality may be modelled, which allows some primitive conclusions to be drawn, such as the 12% increase in chronic illness, and the observation in Van den Berg's paper of the ~5% reduction in lifespan as a result of a recession.