Shutting down the economy and bringing social life to a standstill did not contain the epidemic. The virus spread and in addition to the harm done by the epidemic itself, additional immense damage has come upon many people because of the lockdown.

Problematic predictions

As an economist, I am trained in making forecasts and am thus familiar with the pitfalls of making predictions. Many processes in economics have similar patterns as epidemics. From the spread of new products to recessions and the contagions that happen in currency crises, processes similar to “pandemics” take place. Widely ignored at its inception, and at first, only slowly growing, crises often remain undetected until it is too late. When the problem finally gets attention, the authorities tend to overreact. Countermeasures are taken that do not contain the problem but acerbate the problems.

Often it would have been better to do nothing and let the things run its course. Yet governments are asked to do something. The population gets hyped-up as the media urges uncritically that something must get done. The same type of error as it often happens in economic policy has also been the case with the lockdown in 2020.

Epidemiological prediction models are not better than those of the economists. Not much different from dynamic developments in economics, it is also almost impossible to foresee in time and with clarity the future shape and dimension of an epidemic. Epidemic processes happen in the form of a “growth curve” which is a well-known quantitative development in economics, finance, and biology.

When a process grows at a constant rate, the overall increase starts out small but gets bigger and bigger over time until its unavoidable collapse. A bacteria culture, for example, will grow slowly at first and gain speed over time until it finishes off in a total collapse when no longer any host is left.

As can be seen in the graph below (figure 1), the coronavirus epidemic follows the typical pattern of a so-called logistic growth curve with the characteristics of almost flat growth at the beginning to be followed by an increasingly steep rise until a peak from where the curve bends down.

Figure 1: World-wide daily reported COVID-19 deaths from January 22 to May 15, 2020

Source: Our World in Data. Report May 15, 2020

One can see that the curve is almost flat from January until early March but took a sharp upward turn in the second half of March 2020, first moderately and at the end of March at an increasing speed. From the middle of March until early May 2020, the number of daily confirmed deaths surged until it peaked at the beginning of May and has been falling since the middle of April 2020.

To better explain such a growth process in more detail and to highlight its features, it helps to stylize the curve and concentrate on its beginning and the exponential part of the growth process (Figure 2).

Figure 2: Exponential growth process

The curves in this graph (Figure 2) point out that at the beginning of the process with its move from a to b along the time axis (t), the existence of an exponential process is still unrecognizable. In the real world, under data uncertainty, an early prognosis could easily have been linear as shown by line P1. During the first half of the process from almost flat at the beginning (a-b) to the stretch at which the curve becomes visibly exponential (b-c), half of the time has already gone by (from a to b) until its endpoint (d) has been reached. Nevertheless, around point b, the projection would still suggest a relatively moderate development (P2) as the movement of the curve is still only slightly bent upwards after reaching stage b and the effects of the process (Q1) are still mild compared to what they will be in the end. At the time between b and c, it may seem as if the process is still manageable. After all, the quantitative effect (vertical axis) has only moved up to Q2 on the vertical axis.

The part from c to d represents the dramatic part of the exponential growth curve. In about one-seventh of the total time span from almost flat to almost vertical (at d), the largest part of the effects takes place within the time span from c to d. In this phase, the speed of change is so fast that in the case of a harmful policy issue the authorities get overburdened by the events and fall into panic mood, which promotes taking wrong decisions.

At their beginning, exponential growth processes typically remain undetected. Yet when they are recognized, they often motivate their discoverers to exaggerate their dimension. When it is a policy matter, and the media take up the issue, public decision-making tends to discard sound judgement.

Disastrous decision making has also taken place in the confrontation with the pandemic of the coronavirus. In an attack of panic, governments around the world implemented harsh measures to block the spread of the coronavirus. The lockdown of the economy included the closing of schools, the suspension of sports and cultural events, the closing of restaurants, and the shutdown of nonessential businesses. Many governments implanted a policy of social distancing.

Failed policy of “flattening the curve”