Loading...

Making predictions


Making predictions

In recent times, especially after the Covid-19 pandemic struck the world, predicting about the future trend of the pandemic's behaviour has become a serious issue. Experts in epidemiology are regularly making predictions about the spread of the disease globally as well as in different parts of the world. They predicted about a second wave of the pandemic after its first wave wreaked havoc first in China and then gradually in different parts of the world including Europe and the Americas.

The second wave has really arrived! It is already in Europe and all other countries, including Bangladesh, are awaiting its impact.

So, the epidemiologists' prediction about the pandemic's course has generally been correct. But the financial world and its institutions including the World Bank (WB), the Asian Development Bank (ADB), the International Monetary Fund (IMF) or other such institutions could not prove to be equally lucky in making exact predictions about the prospects of the global or local economies. So they had to review their forecasts about global trade shrinkage as a result of supply chain disruptions following measures adopted to control the pandemic. First it said the global Gross Domestic Product (GDP) will shrink in 2020 by 5.2 per cent, which means it would be the deepest recession that the world has witnessed since the Second World War. According to this forecast, the advanced economies would shrink by 7.0 per cent due to dislocations in supply chain, trade, finance and, in sum, fall in consumer demands. Similarly, the emerging markets and the developing economies would see shrinkage of 2.5 per cent, their first in sixty years!  But now the IMF's chief economist, Gita Gopinath says, "forecasts for the global economy are somewhat 'less dire'. The reason is simple. The rich nations, especially China, have not been as seriously affected as predicted at the beginning and they have bounced back admirably. In its latest forecast the shrinkage is corrected to be at 4.4 per cent from the earlier 5.2 per cent. However, the global economy's return to normality will be weaker than it was predicted earlier in June. Similarly, its earlier (made in April) forecast about Bangladesh's GDP growth in 2020 under the pandemic, it said, could be 2.0 per cent. But its last forecast about the same made on 12 October was that it would be 3.8 per cent. The Bangladesh authorities, since the beginning, have never agreed with global financial institutions' predictions. It appears the government's insistence on the economy's robust performance despite the pandemic is at least partially, if not wholly, correct.

That brings us to a quite different perspective about the art and the science of making predictions.  Predictions are not meant to be accurate. Non-exact sciences like economics are not too fussy about 100 per cent accuracy of their predictions.

For exact sciences, however, there is or, to be more precise, was a problem until the early decades of the 20th century. There was no problem, however, accepting experimental results that were not 100 per cent true to the theoretical predictions. A deviation from theoretically predicted value was acceptable with the qualification that the error was attributable to the crudeness of experiment and the instruments used for the purpose. But ideally, theoretical predictions made by physical sciences should be hundred per cent accurate. But this idea had to be changed in the first quarter of the 20th century.

 In fact, this element of uncertainty in predictions is something built-in not just in experiments. The uncertainty is also in the theory or rather in the reality that theories of sciences try to describe.

Thus making prediction or forecast is a very serious matter for a section of trading people who speculate about the future trends of markets. It is also a tool in the economists' and meteorologists' hand to tell them how economies will fare in the longer and shorter terms or how the weather will behave in different parts of the globe over a period of time. In the exact sciences in earlier times, a measure of a scientific theory's worth was in its ability to replicate experiments and produce results as predicted. In fact, it is still so, but in special cases.

Whether it concerns serious scientific disciplines or fortune tellers who make predictions about a person's or a country's or even the entire world's future, they are basically trying in their own ways to chart movement of events in time. Some practices of predicting such as those in vogue among fortune tellers are often dismissed by the scientific community as bogus. But fortune tellers such as astrologers not unlike the people of science have their own body of knowledge, though ancient, and methods based on that knowledge, to support their art of forecasting. But the astrologers' methods, the people of science would argue, do not have any scientific basis. That means, it is all about the methodologies that make the difference!

But can exact or non-exact scientific disciplines really claim that its method is incontrovertible? Is there not a branch of exact science, as discussed to some extent in the quantum mechanics, that has demolished the very idea of predictability?  Can physicists of modern era say with absolute certainty as could their pre-quantum era counterparts, at least in theory, that a particular physical system would evolve over time exactly in the way they would predict?

Even so, in what they say, at the macro level, predictions may work! So, there are two quite different worlds, the macro and the micro, which cannot see eye to eye!

True, we all live in two different worlds, call it micro and/or macro, at the same time. Common people who, before the sciences could reach their profound conclusions, were happily living simultaneously in two, or rather many worlds! And they had always their wise men, the religious gurus, astrologers and such to turn to for predictions whenever faced with uncertainties about their lives. And they never expected those predictions to be accurate.

 

[email protected]

 

Share if you like

Filter By Topic