Ben's blog and articles

In 1986 the behavioural economists Daniel Kahneman, Jack Knetsch and Richard Thaler published a classic study showing that in a hypothetical snow storm a large majority of people would think it unfair for a hypothetical hardware store to raise the price of a hypothetical snow shovel.

Ordinary people don’t like price gouging, was the finding.

We don’t need hypotheticals to tell us that in the grim age of Covid-19.

Just look at the backlash against Mike Ashley’s attempt to hike the selling price of his weight-lifting equipment at his Sports Direct chain last month, shortly after the lockdown began.

And new Office for National statistics data today shows that the online price of cough and cold medicines jumped by 10 per cent in a week last month, prompting talk of profiteering by digital retailers and demands for a state crackdown.

The reason why the Kahneman paper was well-known was because it was – to economists at least – problematic; another piece of that famous corpus of behavioural research showing that people don’t think and behave according to the axioms of old-fashioned economic textbooks.

Economists generally believe that markets work because prices are allowed to reflect the balance of supply and demand.

So if the demand for snow shovels surges in a snowstorm it’s natural for the price to rise to reflect that - welcome even because it transmits a signal to snow shovel manufacturers to produce more of them. Or because it incentivises others to put their own stocks up for sale.

Prices and price changes are information and if you interfere with the flow of information you’re going to get inefficiency.

What this misses, of course, is context.

People who would, in normal times, be content to see the retail price of a snow shovel fluctuate, perhaps reflecting the costs of production of the manufacturer, feel very differently about the shop owner who whacks up prices in a time of perceived social emergency in order to make a profit - and especially so when it involves goods that are seen as essential such as medicines.

Markets are embedded in societies. And societies – people - care about fairness.

For free market fundamentalists and libertarian commentators to rail against this reality is the economic equivalent of howling at the moon.

There’s clearly a justification for fairness consideration to trump efficiency in this pandemic, certainly when it comes to anything related to health.

Yet the market fundamentalists are not entirely wrong to warn against the popular urge to stamp out price fluctuations.

Wartime-style direction of industrial production works up to a point.

The Government can requisition private businesses to convert their factories to produce ventilators rather than cars. The state can transform conference centres into field hospitals.

But total state control of the economy, even in a health emergency like this one, is neither feasible nor desirable.

It’s generally agreed that we want private businesses to survive this crisis so they can continue to employ workers. But if we hammer their profits margins in the name of fairness we risk undermining this goal.

Happily, it’s likely that most people understand this. One of the perhaps more interesting findings of the Kahneman paper was that a majority of those questioned were willing to accept price rises (and wage cuts) from firms when the firm was in danger of going bust.

Context matters. So how do we apply all that now?

As so often, it’s a balance. Market regulators and the government need to judge when to intervene and when to let the market operate. Businesses need to judge what’s socially acceptable behaviour when it comes to pricing and what isn’t.

It’s won’t be easy. But relative to the other agonising dilemmas thrown up by the coronavirus pandemic it’s not the hardest job either.

The US has now overtaken China in the number of confirmed coronavirus cases. As of Thursday [26 March], America had 85,996 cases, versus 81,894 in China.

There were 17,224 new cases in the US on that day compared to just 56 in China. Meanwhile, China has started easing its own economic lockdown of cities, even Wuhan, where the pandemic originated.

On the face of it, China has conquered the virus, while the US is on a trajectory to be the worst affected country in the world. But is this impression accurate?

There are grounds for caution. First, it's important to remember that these headline statistics that we see are "confirmed" cases, not total cases - an important distinction, says Max Roser of scientific publication Our World in Data. The number of confirmed cases is likely to depend considerably on the level of testing being done in any individual country. Epidemiologists admit that the true number of UK infections is almost certainly many times the number that have been confirmed.

The government's chief scientific adviser Sir Patrick Vallance told MPs on 17 March that as many as 55,000 people in the UK could have the virus. At that stage the UK had fewer than 2,000 confirmed cases from testing.

Germany is now testing 500,000 people a week. By contrast the UK has tested only around 100,000 in total. That may help explain why the Germany fatality rate - the number who die per recorded infection - seems so much lower than some other European countries.

Italy has 80,589 confirmed cases and 8,215 deaths. Germany has 47,373 confirmed cases and 285 deaths.

That gives an Italian death rate of 10 per cent and a German death rate of less than 1 per cent.

It may not be because the virus is somehow less lethal in Germany or because their health service is better at keeping people alive, but simply because Germany is discovering more cases of infection.

There is also a clear correlation between a country's GDP per capita and its number of tests per head of population. That makes comparisons about the prevalence of the virus in different national populations potentially misleading.

It's possible that the disease is more widespread in poorer countries than official reported cases show because they are testing less. Some may be reporting differently too, which brings us to China.

China's National Health Commission says that "asymptomatic" cases of coronavirus infection - where people test positive but do not show clinical symptoms of illness - should not be counted towards its confirmed national total of cases. That's out of line with the practice of the World Health Organisation and also some other countries like South Korea.

Hong Kong's South China Morning Post reports it has seen classified Chinese government documents stating that 43,000 people in China had tested positive for the virus at the end of February but because they were asymptomatic were not counted.

And that could be significant. On Monday the respected Caixin newspaper reported an epidemiologist in Wuhan stating that "every day we can still detect a few, or a few tens, of asymptomatic carriers".

That matters because some epidemiologists think the asymptomatic can still pass on the disease. As the disease expert told Caixin: "We cannot conclude yet that Wuhan's transmission has completely ended."

Some epidemiologists fear that, when the social lockdown is lifted, Wuhan - and China - could be at risk of a second outbreak.

China's success in suppressing the disease is not a fiction. Experts think that number of deaths from the disease being reported by China is a reliable figure, even if the total number of cases may not be.

And that has fallen dramatically, from a peak of 150 on 23 February to just 5 on 26 March. By contrast, on 26 March the US reported 268 deaths.

Yet looking beneath the bonnet of the headline data suggests that declarations of victory against this new virus - not just relating to China but any country - would, at this stage, be dangerously premature.

Oh for a beaker of the scientific controversy of normal times.

Drinking wine’s good for you, says one news headline.

No wait, wine’s actually bad for you, says another.

We ask ourselves: why can’t the confounded scientists just make up their minds rather than giving us such conflicting and confusing advice?

Sometimes the problem is inaccurate or sensationalist reporting.

Sometimes it’s been misleadingly presented or overhyped by the researcher in the press release.

Yet sometimes the appropriate answer is that: science doesn’t actually work like that, particularly in a field like the study of disease, epidemiology.

Different teams of researchers, studying different populations, find different results.

There is no single scientific “mind” to be made up.

Over time, if we’re lucky, we tend to get a clearer picture, because numerous independent studies point in a similar direction. Think of the science of manmade global warming and its ecological impacts.

But science is often a messy and “emergent” process, unfolding over time rather than in a flash of discovery.

Which brings us to the science of Covid-19.

A newly-released study from epidemiologists at Oxford University hypothesises that half of the UK population has actually already been affected by this novel coronavirus, often without knowing it or showing any symptoms.

It’s received a good deal of attention. And understandably. Because if the assumptions of this study are true the implications would be profound.

It would mean that the fatality rate – the proportion of people likely to die from contracting the virus – is significantly lower than the range official experts have hitherto been assuming.

It would also mean that we could look forward to a much earlier lifting of the social lockdown than otherwise.

It would mean that we have already built up a degree of herd immunity to the virus, making the threat of being hit by another wave of disease later in the year significantly lower.

It would, in short, be good news.

But is it true? We simply can’t know. But it’s worth probing, as the Oxford study authors say, through large-scale blood testing.

If significant numbers of asymptomatic people turn out to have antibodies against this virus that will give statistical support the hypothesis.

Boris Johnson says that throughout this crisis his government has been “led by the science”.

The claim that the science changed suddenly last week – when the Government suddenly shifted from a mitigation to a suppression strategy for the virus – has been met with derision in some quarters.

The science has long been crystal clear, insist critics. Spiralling infection and hospitalisation rates in Italy were obvious. The government should have ordered a total lock down of the UK economy and society weeks ago.

Perhaps it should have done.

Yet the characterisation of “the science” as a fixed thing here isn’t helpful and could even be dangerous.

There is no science as such, rather an array of scientific evidence which requires judgement by experts and policymakers to interpret. And the reality is that in a pandemic like this one new scientific information – about fatality, recovery and transmission rates – is constantly emerging.

New information about total infections could change the whole game.

It’s important to note that epidemiological projections of the spread of a disease requires computer modelling. And experts can differ on the appropriate parameters, or inputs, to those models.

By all means question the judgements made by politicians – and even experts like the chief medical officer and the Government’s chief scientific advisor – about the evidence available. Ask whether the parameters of the modelling being used are appropriate. Ask whether the models are up to date given the new information continually emerging.

Indeed, it’s vital that is done by other independent experts, and regularly, to counter the perils of “groupthink” among policymakers.

But let all of us, also, try to be a bit more sophisticated about the nature of “science” in a situation like this.