Reuters BreakingViews title: Risky business
Author: Edward Chancellor
Date: 5 June 2020
“We live in a world in which policymakers must grapple with great and novel risks. The trouble is they keep making a hash of things. At the turn of this century, the Federal Reserve acted to ward off the threat of deflation, while at the same time the George W. Bush administration was seeking to protect Americans from Saddam Hussein’s weapons of mass destruction. While those turned out not to exist, the Iraq war destabilised the Middle East at a huge cost in lives and treasure. As for the Fed, its easy-money policy fuelled a housing bubble which sparked the global financial crisis. There’s little reason to believe that policymakers are handling the coronavirus outbreak any better.
The late German sociologist Ulrich Beck coined the phrase “risk society” to describe the modern world’s leading obsession. For Beck, risk is an anticipated danger – not something that could easily be measured in the way economists define it. A new notion, known as the “precautionary principle”, dictated that preventative measures should be taken against large-scale risks, however unlikely they might be. Beck warned that politicians were likely to overreact: “It is not going to be easy in future,” he predicted, “given the state’s promise of security and a mass-media hungry for catastrophe, to prevent a diabolical power game with the hysteria of non-knowing.” That sounds like a pretty good description of life under Covid-19.
So how should leaders respond to the challenges presented by the pandemic? In their recently published book, “Radical Uncertainty”, John Kay and former Bank of England Governor Mervyn King describe decision-making for an unknowable future. To summarise the message of this profound 528-page tome: it’s not rocket science that’s required, but common sense. Complex mathematical models aren’t useful guides to action due to a paucity of accurate data. There are too many of what Bush’s Secretary of Defense Donald Rumsfeld called “unknown unknowns”. In an uncertain world, the greatest challenges are unique. Furthermore, human behaviour is constantly changing in response to new opportunities and threats. In technical terms, social data are non-stationary.
This means that we don’t have accurate probability distributions upon which to base decisions. But that doesn’t stop people acting as if we had them, especially in the world of finance, where number-crunching is addictive. At the onset of the global financial crisis, the chief financial officer of Goldman Sachs, David Viniar, notoriously referred to the markets experiencing a 25-standard deviation event. What he meant was not that something impossibly rare had occurred, but rather that Goldman’s risk models couldn’t comprehend what was going on. Kay and King describe the “Viniar problem” as “the mistake of believing you have more knowledge than you do about the real world from the application of conclusions from artificial models.”
The current approach to the Covid-19 outbreak has been another example of the Viniar problem in action. Just as Wall Street had its financial risk modellers, politicians have become excessively dependent on mathematical epidemiologists. The high death-rate forecast produced by modellers at Imperial College London is said to have prompted the British and American governments to adopt a lockdown. Since that date, there has been much discussion of the lowering of the virus’s rate of infection or reproduction number, a key input in the model. Yet the R number is inherently unknowable, at least in real time. Nobody knows how much of the population is already immune to Covid-19. And the spread of the disease has changed people’s normal behavior, which means the data are non-stationary.
Kay and King suggest that when faced with uncertainty the best response is to ask a simple question: what is going on here? Rather than arrive immediately at a fixed answer, remain open-minded. Wise decision-makers resemble foxes rather than hedgehogs – “the fox knows many things, but the hedgehog knows one big thing,” in Isaiah Berlin’s famous riff on an idea first attributed to ancient Greek poet Archilochus. The reference narrative – what’s going on here? – can be adapted and changed as new information becomes available. While models can help to frame problems and examine scenarios, decision-makers should never use a single one. That’s a route to groupthink and eventual failure. Instead, they should employ a plurality of models and wait to see which comes up with the best results. The exercise of judgement is unavoidable.
Since the coronavirus pandemic, however, common sense has played second fiddle to mathematical models. Over time, it has become evident that Covid-19 has nosocomial traits, meaning its contraction and fatalities are largely confined to hospitals and care homes. Had policymakers not been so obsessed with the models, would so many infected patients, on both sides of the Atlantic, have been moved from hospitals into nursing homes with such tragic results? Instead of using multiple models, single models – such as Imperial’s – have become dominant. Groupthink prevails. Despite their immense fiscal, economic and social costs, there has been no open and honest appraisal of the efficacy of lockdowns. Only the Norwegian health ministry has openly suggested that its own lockdown may not have saved many lives. Yet unless a frank assessment is conducted, it will be impossible to know whether to continue with the lockdown policy in the event of a second outbreak.
Beck suggested that the modern world faces a clash of risk cultures, rather than of civilisations. He was right. Witness the daily bickering over the efficacy of face masks, appropriate social distancing and whether schools and businesses should remain closed, with the battle lines drawn up along existing political divisions. It would have been encouraging to see politicians rise above the fray, leading their peoples through a difficult period of uncertainty and fear. Instead, they have used epidemiologists’ models – “following the science” – to shield themselves from personal responsibility for their decisions. As Kay and King observe, evidence-based policy has become policy-based evidence. If “risk societies” are to prosper, they must improve their response to the perceived dangers thrown up by an uncertain world. Otherwise, we really are doomed.”
——
Source:
https://www.breakingviews.com/columns/chancellor-a-catastrophic-failure-of-risk-control/
0 Comments