Department of Agricultural, Food, and Resource Economics Food Security Group

COVID-19 is a monstrous challenge for the development community not just because of the enormity of its potential impact, but also the enormous uncertainty around what that impact might be.  Nowhere is this better captured than in the latest estimates of Infection Fatality Ratios (IFRs).  IFRs are what we need to properly respond to a pandemic such as this, but they are almost never reported in early stages because the denominator – the number of people infected, not the number of confirmed cases as in widely reported Case Fatality Ratios (CFRs) – is subject to so much uncertainty.  What is certain is that IFRs, once they are known with more precision after a pandemic has run its course, are always lower than CFRs – typically much lower. 

Credible IFR estimates have now begun to emerge, five months into the pandemic, but their range is enormous, from 0.1% to sixteen times that figure, 1.6% (Table 1).  The lowest figure of 6.9 million estimated deaths is, relative to the current global population of 7.9 billion, less than one-tenth of the lowest estimate of deaths from the 1918 Spanish flu, while the top estimate exceeds the lower-bound estimate of 20 million global deaths from that pandemic. 

Table 1: Recent estimates of infection fatality rates (IFRs) from COVID-19, and implied global deaths, assuming no mitigation

Notes: Riou et. al. use the term CFR rather than IFR, but refer specifically to death rates among infected individuals; Riou et. al. make no claim that their estimate is applicable on a global scale. We use it here to illustrate the wide range of credible, scientifically generated fatality rates associated with COVID-19; Walker et. al. IFR estimate is computed by author based on global infection and death figures; CEBM IFR estimates are directly reported.

These differences matter as African governments think about how to confront the pandemic in their countries.  Economic losses of the magnitude likely to emerge from our global response to this pandemic also imply large loss of lives and of quality of life.  These losses occur over a much longer period of time and are unquantifiable except through modeling, so they are frequently ignored in the early stages of decision making. But we know they will occur. 

We also know that the human consequences of these economic losses will be dramatically more severe in poor countries, for at least two reasons: hundreds of millions of people already live far too close to the edge of disaster in their every-day lives and any shock can push them over the edge; and governments of these countries have nowhere near the capacity to replace lost incomes that high income countries do (see our first blog post). Deep economic downturns generated by lockdowns will have devastating effects on the livelihoods of hundreds of millions of people, will leave many of them indefinitely trapped in even deeper poverty than they now experience, and this will cost lives. 

A credible, immediate commitment by the global community to replace large shares of the incomes of affected citizens in low income countries could make this line of thinking unnecessary, or at least less necessary, but current discussions do not come close to matching the need.  As a point of comparison, the US government has approved spending USD2.2 trillion to address the economic fallout of COVID-19 and may soon approve another USD1 trillion – a total of USD 3.2 trilllion. This is about USD10,000 per capita, or roughly 16% of US GDP during a year.  Total GDP in LICs and LMICs is roughly USD7.3 trillion, meaning that a similar level of support for these countries would cost nearly USD1.2 trillion. The global community has, of course, not come close to proposing such a sum. 

In the absence of a global commitment of something approaching this amount, we must ask the question: how much economic pain is too much as we combat COVID-19 in the developing world?  None of us know the answer, but we must, insistently, ask the question.