On the eve of the UEFA European Football Championship, a BBC reporter asked an English football fan which side he expected to win.
The reporter prefaced his question by saying that “I know predictions are meaningless today, but I’m going to ask you anyway.” The fan then predicted that the English team would win because it was playing on home turf, had the better line-up, and had performed better than its rival. It was the “team of hope” that the English had been looking forward to for over half a century, and it would crown them with victory, he said.
He went on to say that he and his friends planned to celebrate the following day, which the government had made a public holiday. Of course, as we know, when the match was played last Sunday England and Italy tied on goals, and then the latter won on penalties. So, it was Rome that celebrated while London mourned, and the English fan and his friends probably spent their day off debating whether their team’s loss was due to rotten luck or a fault by this or that player.
This disappointment is an example of how difficult it is to make predictions. As the old joke goes, predicting is one of the hardest things you can do, especially when speaking about the future. This is because by its very nature, forecasting is a prisoner of past experience, and this cannot necessarily be used as a guide to the future.
Decisions affecting the future are limited by the knowledge that is available to us in the present. They are also affected by bias, as was the case with the football fan above. Expectations can be misled by various types of “noise” that distort clear thinking.
In their recent book “Noise: A Flaw in Human Judgment”, Nobel Prize Laureate in economics Daniel Kahneman and co-authors Olivier Sibony and Cass Sunstein discuss bias and noise as main components of human error. The former they define as a “systemic deviation” formed in advance, as is the case with prejudice. For example, if a judge investigating a crime believes that a certain group in society is more prone to crime than others, that prejudice could dominate his thought processes if a member of it is brought before him as a suspect, even blinding him to the facts.
As a result, truth and justice are not served. Kahneman and his colleagues maintain that such biases are well known and that they can be counteracted by various means in order to arrive at the truth. In the example above, a suspect wrongfully sentenced as a result of bias can challenge the ruling on appeal on the grounds of due process and evidence of the judge’s wrongdoing.
Noise, which Kahneman and his colleagues describe as “random scatter,” is a no less serious and more intractable problem than bias. In the example of the above-mentioned judge, bias for or against a particular group may only be part of the problem. Even in two very similar cases involving suspects from the same group, the judge might deliver two strikingly different rulings depending on how his emotional or physical health was at the time.
In short, when issuing a legal ruling, making a medical diagnosis, or any other such decision-making process, it is important to minimise the effects of both components of error, bias and noise, in order to ensure that decisions are as sound or as “distortion free” as possible.
In economics, the higher the risks, the tougher it is to make decisions and the more crucial it is to eliminate bias and noise. As a result, economists formulate probability scenarios as a way to take decisions as rationally as possible. But when conditions are too uncertain to forecast, such formulations become futile and decision-making becomes akin to shooting at a target in the dark. In such cases, the costs of decision-making and pre-empting any adverse effects skyrocket.
As we contemplate the results of the G20 group meetings that concluded last week, we find a sterling example of the forecasting problem mentioned above. Their resolutions rested on the prediction of an economic recovery this year and next and the forecast that growth would range between 5.5 and 6.5 per cent.
But that forecast is predicated on predictions concerning the recovery from the Covid-19 pandemic and the ability to stop the spread of a virus which is still taking enormous tolls on human life and standards of living. These in turn are predicated on the efficacy of the vaccines against the virus and its variants, and on how effective the economic-stimulus measures will prove to be in boosting production and creating jobs.
Four main problems already hamper such economic predictions, which are intimately connected to interrelated assumptions.
First, there are insufficient amounts of vaccines and a lack of availability in developing countries, raising the spectre of new waves of infections and deaths, overwhelmed health sectors as the virus mutates, and the transmission of variants across borders, including to developed nations that have managed to inoculate large portions of their populations against the current variants.
Recall that the G7 group’s pledge to send a billion vaccines to the Developing Economies covers less than 10 per cent of these countries’ deficits. Meanwhile, there is still strong resistance to allowing the vaccines to be produced in developing nations due to the developed nations’ insistence on upholding the intellectual property rights of the firms producing the vaccines despite the developing countries appeal made at World Trade Organisation for a temporary suspension of these rights.
Second, there are disparities in the financial capacities to combat the virus and promote recovery. While the developed nations have spent over 25 per cent of their GDP towards these ends, the ratio plunges to seven per cent in countries with medium-size economies and less than two per cent in the lowest-income countries.
Third, there are the repercussions of inflation, especially in the US where the annual rise in consumer prices hit a record of five per cent in May. This relatively high rate for the developed nations will have an impact on interest rates, the cost of loans and foreign debt stability, not to mention the impacts of the rising costs of commodities on the budgets of countries compelled to import them.
Fourth, the above-mentioned growth prediction is not linked to job creation, which translates into more acute disparities in income and wealth distribution.
These details are important when analysing economic-growth forecasts, because it is in such details that the devil lurks. So far, the recovery has been reduced to certain fields, which means that recovery measures and their results will not be comprehensive, stable or sustainable. What we are watching, in effect, is growth without jobs and gain without pain.
The figures mentioned are not linked to measures to put some 250 million unemployed people back to work. The number of billionaires is increasing, and their wealth is growing while more than 120 million people are being pushed into abject poverty.
The international disparities in recovery rates can only bring more tension, conflict and social upheaval, which always follow crises. The current crisis is the worst that has been seen in the contemporary period. It demands firmer and more urgent measures to be taken to save people’s lives, protect their livelihoods and safeguard the stability and security of their societies.
*An Arabic version of this article appeared on Wednesday in Asharq Al-Awsat.
*A version of this article appears in print in the 8 July, 2021 edition of Al-Ahram Weekly