Thoughts from Pontevedra, Galicia, Spain: 2.5.21

Night’s candles are burnt out, and jocund day stands tiptoe on the misty mountain tops

Spanish life is not always likeable but it is compellingly loveable

– Christopher Howse: ‘A Pilgrim in Spain’ 

NOTEInfo on Galicia here. Detailed info on Pontevedra coming (relatively) soon.


Spain: A questionable development?

Moving on: The three things you’ll need to make sense of a changing world. See the excellent article below.

Cosas de España

Lenox Napier on speaking Spanish.

Mark Stücklin here again addresses the issue of squatters: In Spain, he says, there is a greater risk of squatters getting in than in other countries, and once they are in, the cost to owners is higher than in other countries. Put another way, there’s a bigger chance of it happening, and when it does happen, it’s much worse.

Cousas de Galiza

A propos . .  The Voz de Galicia had an article on 16 April with the headline: The Xunta warns of an acceleration in Squatting. And then adds: The Xunta calls for [national] legal reform to end the incongruence of greater penalties for robbing a phone than taking over a property. There’s been at least another 6 articles on the issue since then in that paper alone. So, yes, a pretty serious issue. 

Entrepreneurial local travel agencies are offering trips to Russia to those who don’t want to wait for one of the variants available here and are happy to have the Sputnik option.

Plans for Pontevedra: I looked up ‘intelligent tourism’. It comes under the rubric of Alternative tourism.

María’s Level Ground: Day 28

The UK

Boris Johnson’s salary is £157,372 pa. This is ‘close to the top 1% of UK earners’ but leaves him 13th in the list of 20 international peers, with Germany and Sweden above him but France and the Netherlands below him. Top of the list is Singapore(£1,592,299) and the bottom 2 are Spain (£72,188) and Brazil(£53,517). I never fancied being PM here anyway.

The Way of the World 

In case you need to know, this is what transhumanism is.


This sentence appeared in The Times today: Senior Tories say the affair has shined a light on Johnson’s financial difficulties. I’d never heard ‘shined’ before, only the irregular ‘shone’. But both exist, it says here. Perhaps frequent ‘wrong’ usage has introduced the regular version. As ever, the people decided, not an academy.

Finally  . . .

I checked it out and it’s true . . . St Francis of Assisi really did aver that fleas and other bodily parasites are the pearls of poverty, as gifts from God. Which is as good a reason as any to discount anything else he postulated. But no doubt it helped to think this during the Black Death.


The three things you’ll need to make sense of a changing world: Tim Harford, The Times

There is a famous legend about Galileo’s telescope. The tale, which has grown in the telling, is that when the great Italian scientist was being prosecuted by the Catholic church for professing that the Earth and planets orbited the sun, he invited the cardinals to take a look through his telescope and see for themselves. The cardinals refused to look, claiming that the telescope itself was some sort of trick.

The story tempts us to feel satisfied with our modern rationality but we shouldn’t be so smug. Far too many people refuse to gaze through the lens of statistics today. We, too, are afraid of being tricked.

There is a reason for that fear. Margaret Thatcher and John Major endlessly redefined “unemployment” in order to reduce the headline figure. Tony Blair and Gordon Brown were masters of statistical misdirection. The Leave campaign boldly lied about the numbers; indeed, their aim was to be accused of lying, with a noisy row playing to their strengths.

No wonder we have been reluctant to look through the statistical telescope. Britain sent £350 million a week to the EU? Brexit would cost 820,000 jobs? Whatever number those liars on the other team were saying, nobody believed a digit of it. Whether or not the British people had had enough of experts, we had certainly had enough of statistics.

Then came Covid and everything changed. Amid the tragedy there was a striking change of tone. Our knee-jerk cynicism was quietened. We had questions: how dangerous was this virus? How did it spread? Who was most at risk? Suddenly, we needed the numbers.

For a proud geek like me, it was refreshing that statistics were no longer being used as a weapon but as a tool. Like radar, ultrasound, or the humble telescope, data show us things that are otherwise invisible. But they can only help us if we use them wisely; indeed, they can only help us if we are willing to use them at all. So what has the statistical telescope shown us recently, where are its lenses cracked or warped and, most importantly, how can we use it to bring the world into focus?

Step one: keep calm

If the past 5 years have taught us anything, it’s that what we believe or disbelieve is as much to do with our hearts as with our heads. We tell ourselves that we rely on the facts but our preconceptions and our politics influence what we believe the facts to be. So at the risk of sounding like a mindfulness guru with a calculator, I believe we should all notice our emotional reactions to the statistical claims that swirl around us. The media and social media thrive on fear, joy, anger and smug vindication. There’s nothing wrong with feeling emotions but we are not at our wisest when rushing to rage-tweet about a claim we did not actually check. So take a moment to notice your instinctive reaction to that astounding piece of data. Then look again.

Eric Feigl-Ding should have done just that in March last year. Dr Feigl-Ding, an American epidemiologist with a large Twitter following and an excitable style, shared a graph from the US Centers for Disease Control and Prevention adding, “Newsflash for YOUNG PEOPLE: You are not invincible! You’re just as likely to be hospitalized as older generations — even CDC says so!” The CDC graph actually said the opposite, albeit in a confusing graph. It showed that vastly more over-45s than under-45s were in hospital. Dr Feigl-Ding should have looked at the small print, otherwise known as the x-axis.

A few days later the right-wing US pundit Ann Coulter tweeted a pair of graphs with the comment “For people under 60, coronavirus is LESS dangerous than the seasonal flu.” Again, the graphs showed the opposite. But Coulter’s “black is white” message was retweeted more than 10,000 times. When passions are running high and the situation is urgent, who has time to stop and think?

If you’re not experiencing any emotional reaction that, too, is worth noticing. In February last year the infectious disease expert Dr Nathalie MacDermott told me that the new coronavirus seemed to kill nearly one in 100 infected people and might easily infect two thirds of the world’s population. These were shocking numbers but the first was accurate and the second would probably have happened without dramatic lockdowns. We were looking either at 50 million deaths or a radical change to our way of life.

When I look back on that conversation, I’m struck by how little emotion it provoked. I could do the maths but I couldn’t grasp what it might mean. I later discovered that psychologists have a term for this: “negative panic”. Sometimes the emotion you don’t feel is the emotion you should be noticing.

Step two: get context

Statistics can seem bewildering but you can get a long way with simple questions about context. What is being measured here? Is it going up or going down? Is it big or small?

Consider the recent alarm over the possibility that the Oxford-AstraZeneca vaccine might, in rare cases, cause a dangerous form of blood clot. How worried should we be? The UK regulators have published data suggesting the risk of a fatal clot might be two in a million. That doesn’t help much without context. “Two in a million” sounds reassuring but “risk of fatal blood clots” does not. So get context. Find a useful comparison. What else carries a two in a million risk of death?

The answer, according to the Cambridge University statisticians Sir David Spiegelhalter and Mike Pearson, is a 500-mile drive in a car, 40-mile cycle ride, or a 12-mile trip on a motorcycle. Ride a motorbike to the vaccination centre and your journey is more dangerous than the side effects of the vaccine.

Scientists are still trying to quantify the exact risk of fatal side effects: some European regulators think they are higher than two in a million. I will admit that such comparisons can be treacherous. Last summer I compared the daily risk of catching a fatal case of Covid to the risk of dying in the bath. The claim went viral — so vivid! — but I was wrong, not about the risks of Covid, but the risks of bathing. Despite my statistical gaffe, the basic principle is sound: when confronted with a number that you can’t visualise, compare it with a number that you can.

Statistics are often presented without any context at all: in a political speech, in a headline, or in a social media meme. Context-free claims should be a red flag, showing that all is not well. Fortunately, Google giveth what Facebook taketh away: it has never been easier to get the bigger picture with a couple of clicks to a reliable source. For example, the Canadian prime minister Justin Trudeau said last month that the UK was “facing a very serious third wave”. Thirty seconds on a site such as Our World In Data would have showed that our official case count was much lower than Canada’s and falling. 

Any of us can now fact-check a prime minister in seconds from the comfort of our armchair.

Another example: blood-curdling headlines such as “Moderna admits South African Covid variant reduces its vaccine antibody levels SIX-FOLD”. Six-fold? IN CAPITALS? Holy escaped variants! While the six-fold number is both terrifying and true, it is irrelevant. The vaccine still seems to protect against the variant just as well as against the original virus.

The most context-free of all statistical claims is the one that uses no statistics at all: the dramatic photographs of groaning shopping trolleys before the first lockdown, alongside reports of panic buying.

The statistics could have brought much-needed context. Kantar, a consultancy, found that in mid-March last year only 3 per cent of shoppers had “bought extraordinary quantities” of pasta. This wasn’t a story about hoarding but people picking up a little extra, and about the fact that more people were visiting the supermarket instead of the office canteen.

That one statistic from Kantar could have reassured us that the shelves would soon be full again and, perhaps more importantly, reassured us that self-centred hoarders are a tiny minority. However, it was not widely reported: underwhelming, vaguely reassuring statistics rarely are.

Step 3: be curious

Vaccines can be political. The data already suggest a gap emerging in the US, where Democrats are getting vaccinated, while Republicans are more likely to hesitate or refuse. It is a striking example of how polarisation can influence our thinking, even on life-or-death matters that should have nothing to do with politics. Researchers at the Cultural Cognition Project at Yale University study this sort of stubborn tribalism and have found an unexpected cure: curiosity. Curiosity means being open to surprises and willing to admit to gaps in your knowledge so perhaps we should not be too shocked that more curious people are less politically tribal in their reasoning. Most of us tend to be dismissive of new information that challenges our preconceptions but curious people tend to be more open-minded, finding the new facts intriguing rather than threatening.

I’ve found it fascinating to learn more about vaccines. Consider Pfizer’s first earth-shaking statistic: its vaccine was more than 90 per cent effective. Behind that number were the stories of 43,538 trial volunteers, along with the scientists at BioNTech who developed the vaccine. Who knew there were so many ways to create vaccines? Who knew the manufacturing process was so complex? Who knew that they make vast quantities of glass vials in the Venice region, which I had imagined specialised only in the finest and most decorative glassware? The 90 per cent number is crucial but only the beginning of the questions.

By the time the long-awaited needle slid into my arm last week I didn’t have the mental space to be anxious: I was too busy being curious.

Fixing cracks in the statistical telescope

This three-pronged approach — calm, context and curiosity — will get you a surprisingly long way in making sense of the numbers without the need to resort to a spreadsheet. Of course, no matter how much calm, context and curiosity we bring to the numbers, the data themselves need to be reliable. The pandemic has also shown us that there are some cracks in the lens of our statistical telescope.

The most obvious was the lack of testing capacity in the early weeks of the pandemic. The government started making bold promises about scaling up testing, then misled the public by claiming to have hit the targets. The signature move was to claim that a test had been performed the moment a home-testing kit was posted out, despite the fact that it later transpired many tests were never returned. The subterfuge was shameful and the early lack of tests cost lives.

A second statistical pratfall came in October, when it emerged that more than 15,000 positive test results had dropped out of the test-and-trace system because an outdated version of Microsoft Excel (not enough rows!) had been used somewhere in the data pipeline. Amusing as that error might seem, it also cost lives. It is hard to think of a more telling example of the way that we suffer when our statistical infrastructure is creaking.

We are not the only country to have problems. Last year US health departments found themselves sharing outbreak data by fax, while volunteers had to step in to gather data about testing because the federal government didn’t have it. Europe’s economic

statistics have creaked under the strain of reflecting the impact of schools shifting to remote learning, or hospitals cancelling routine procedures. The UK’s statistics are probably more accurate which, alas, means the story looks worse.

These failings should not blind us to the progress that has been made. The scale of daily testing is now vast. Daily data on the pandemic — deaths, hospitalisations, cases and vaccinations — are available to you and me on an easily accessible dashboard at the same time that they are available to the prime minister. The Office for National Statistics conducts a large biological survey of the country, allowing it to estimate the prevalence both of the virus and of antibodies. The NHS has led the world in running rigorous clinical trials for Covid treatments: the Recovery trial, organised by researchers at Oxford University, is open to any Covid patient in any NHS hospital in the country. It has discovered a treatment, dexamethasone, that has already saved a million lives around the world.

All this should stiffen our statistical sinews and catalyse our calculators: the numbers matter. When the boffins and the policy wonks collect them and analyse them in the right way, they save lives. To help the rest of us understand an invisible threat in a confusing world, we could do far worse than turn to the lens of statistics. We can’t all pick apart the technical details but we can keep calm, get context, and be curious. That statistical telescope isn’t a club to beat up your opponents, or a prop for your wobbly arguments. It’s designed to bring things into focus. So look through it and gaze around.