Sometimes breakthroughs come from unexpected sources. On 3 April 2020, Italy hit a sad low, with 1,000 COVID-19 deaths in one day. Late in the evening, however, Professor Bart Van Looy and his PhD student Kristof Decock could reassure their Italian friends and colleagues that after the weekend, the worst would be over.
How could two economists make such a statement, at a time when no epidemiologist in the world could predict the course of the pandemic? For Van Looy and Decock, both affiliated with the Faculty of Economics and Business and Flanders Business School, the answer lies in the ‘scenario theory’ that they study.
Far from the world of viruses, they had developed models to predict how quickly new technologies will spread in the market in the short and medium term. For example, in 2016 they predicted how many electric cars would be sold in Europe in 2018, with an accuracy rate of no less than 98.23%.
In mid-March, Van Looy and Decock realised that their models could also prove their usefulness in the corona crisis. They joined forces with epidemiologists from the Institute for the Future and the Rega Institute. And so they succeeded in predicting both the peak in deaths and admissions to intensive care, for our country and for several other countries.
They did this by first calculating all possible scenarios, says Van Looy. ‘There could be literally hundreds of thousands. We then looked at which scenarios most closely matched the unfolding reality. And so we ended up with a limited number of possible scenarios. These resulted in very reliable short-term forecasts, including the timing of the peaks.’
Van Looy and Decock look back with great satisfaction on those hectic two months in which they worked day and night. ‘The feeling that as an economist you can contribute, that you’ve been able to bring hope to others, is worth its weight in gold.’
Moreover, they have taken giant steps forward in their research. ‘Two months of COVID-19 meant sixty times more new data. It takes years to draw the same conclusions from car sales figures. And if it also turns out that you are introducing a new way of modelling, that you’re solving a problem that people have been struggling with for decades, then yes, you’re going to be very content.’