by David Boyle
The big problem is what numbers won’t tell you. They won’t interpret. They won’t inspire and they won’t tell you what causes what. Statistics have nothing to do with causation, the pioneering number-cruncher William Farr told Florence Nightingale in 1861: ‘You complain that your report would be dry. The dryer the better. Statistics should be the dryest of all reading.’ But over-reliance on numbers sweeps away your intuition along with ideology. It leaves policy-makers staring at screeds of figures, completely flummoxed by them, unable to use their common sense to interpret the babble of competing causes and effects – unable to tell one from the other.
If men with long ring fingers are subject to depression – as they are for some reason – that might alert you to looking for a causal link. The same is true of other peculiar numerical links: high stress makes you much more likely to catch colds, accident rates among children double when their mothers are miserable. These odd connections might surprise and inspire you to think about problems in new ways, but it won’t tell you what causes what. You will have to use your intuition to work out where to look in a massively complex world of complex systems. ‘Scientists try to avoid emotions and intuition,’ says the British biologist Stephen Hardin, ‘but it is exactly those that give them ideas.’
Too many numbers also drives out history – it gives us no sense of the different ways in which people measured in the past. It drives out creativity, locking away Keynes’ dark woolly monster of ideas. And it drives out morality too – leaving our poor beleaguered ethics committees desperately trying to measure themselves a coherent attitude to the frightening future of genetically-modified human beings. And to get through the next few perilous decades, to look after each other, and solve the looming problems ahead, we’re going to need all the judgement, intuition, history, creativity and morality we can possibly muster. So we have to make absolutely sure our tidal wave of measurements doesn’t drive those things out.
What can we do to make sure they survive? We could try measuring more and we could try measuring less. In fact, we can probably do both.
Measuring more is what many of the pioneers in this book have been doing. John Vasconcellos measuring the success of schools and prisons by their ability to give people self-esteem. Simon Zadek and the social auditors measuring corporate ethics. Perry Walker and Hazel Henderson measuring salmon, frogs, vegetarian restaurants and everything else. Measuring more destroys the tyranny of the bottom line. It undermines the great importance of the big number. It punctures the pomposity of the men in white coats and the men in grey suits.
We can do more than that. We can encourage people to measure locally what they think is important, not what they’re told to measure. We can de-standardize, get the subjects of measurement to do their own measuring – the pupils, the patients, the poor.
But we could also try measuring less. This is a trusting, conservative approach. Good professional doctors or development economists will tell you that they can know very quickly what is wrong with the patient or economy – but they then have to spend a great deal of public money collecting the figures in order to persuade anybody else. Measuring less saves money. It also requires considerable faith in other people, and that’s in very short supply these days. It means giving more hands-on experience to schoolchildren, managers, civil servants, police. It means lecturing less and listening more. It means decentralizing power. Most of all it means practising using our imagination and intuition.
Measuring more is the trendy radical solution, but my heart is probably in the opposite. Measuring less means trusting enough to lay aside occasionally some of those systems or policies which ensure we make appointments without discrimination, or decisions without ignoring the people who will be affected by them. Those are worthy reasons for counting, but we die a little if we do nothing but count. A world where we count more is stricter and fairer, but it has less life than the world where we count less. When we count less and get it wrong, we risk inefficiency, bigotry, ignorance and disaster. But when we count less and get it right, we probably get closer to joy and humanity than we can any other way. Human beings can deal with a complex world better than any system or series of measurements.
IV
‘There were once eleven generals who had to decide whether to attack or retreat in a battle,’ wrote Carl Jung’s interpreter Marie-Louise von Franz, introducing the ancient Chinese tale which changed her understanding of numbers. The generals got together and had a long debate, at the end of which they took a vote. Three wanted to attack and eight wanted to retreat – so they attacked. Why? Because three is the number of unanimity.
The story was so shocking, she said, that it woke her up. Suddenly she understood that the Eastern view of numbers was different from the Western one. For us modern Westerners, numbers can only count. Our numbers mount up to cumulative totals – things get bigger, we demand more. In the East, they have significance, meaning and quality. Since Pythagoras’ day, it’s been hard to bridge the gap between the two, but – with Western figures driving out our sense of such things – we may be at a moment in history when we have to try.
When our decision-makers seem to be able to do nothing but count, you can see in reaction a new longing for significance and complex truth, for poetry – despite Bentham’s denial of its existence – and for the sacred. People want rhythms and music rather than bald statistics. We don’t want data, we want enlightenment. We don’t want numbers, we want meaning. And most of all, we need to stop muddling them all up, to realize where the former stops being helpful in the endless search for the latter. We need to help people disentangle from numbers and connect with the kind of understanding that can help them change – because change, and whether it is possible, politically or personally, is the key issue for our generation. Change without numbers is impossible; change without anything else is impossible too, and we need some Universities of ‘Unlearning’ to help us find it. In short, we need a Campaign for Real Wisdom.
This is how Prince Charles made the point in his millennium broadcast on the BBC:
Two and a half thousand years ago, Plato was at pains to explain through the words of Timaeus that the great gift of human rationality should not be disparaged. Far from it, he said – it should be exercised to its utmost, but it must not make the mistake of believing it has no limits.
The same is true of counting. Western numbers that split things up, that see only the parts, which are blind to the most important things in life, can only get us so far.
But in case this doesn’t convince you, I’ll end with an old Scottish proverb which seems to put this point in just nine words, two verbs and thirteen vowels: ‘You don’t make sheep any fatter by weighing them.’
Bizarre units of measurement No. 12
Foot pound
(A unit of labouring force, applied equally to machines and people, by the engineer William Whewell in the 1840s. A working man could lift water at the rate of 1,700 foot-pounds per minute for eight hours a day.)
* * *
Average time British people spend on hold on the phone every year: 45 hours
Proportion of people from India who have not heard of the USA: about 30 per cent
Postscript
We recognize that, in the past, testing has helped reduce unfairness in allocating opportunities and directing resources to the economically disadvantaged, and has been useful for making decisions. However, the growing over-reliance on testing over the past several decades deprives the nation of all the talent it needs and sometimes conflicts with the nation’s ideals of fairness and equal opportunity.
US National Commission on Testing and Public Policy, 1998
Just as I had finished writing the hardback edition of this book, and copies were running off the printer, there was suddenly one of those once-in-a-blue-moon moments of global revelation about counting. It was all a matter of votes. Voting is a bit like a jury decision: we agree to accept a numerical verdict, not because it means someone is objective
ly innocent or because they are objectively the best potential president, but because there are no other ways of measuring such things. The problems happen when the numbers don’t work either.
That’s exactly what happened in Florida during the November 2000 US presidential election. Despite a political season where competing statistics were flung over the networks with ever more abandon, it suddenly became clear just how muddling, subjective and confused by human error the business of counting and measurement is. Far from being an objective assessment of the will of the people, you could count Florida’s votes in any one of a score of different ways.
Like other kinds of statistics in modern life, votes are supposed to be a pretty exact way of finding a new democratic leader. But, if you insist on them being absolutely precise, the whole edifice starts to unravel. And when the winner hangs on 300, 327, or 900 votes out of the 6 million plus from Florida, then every vote counts. But, as the pundits and politicians soon discovered, whether or not a vote was really a vote might depend on how far the ballot had been punched through. It was the constant re-counting in some of Florida’s counties that put the little word ‘chad’ – the tiny bits of cardboard normally knocked out of a computerized ballot card – firmly into the English language lexicon.
Try as they might to recreate laboratory conditions at the re-counts, it just couldn’t work. There were rules in Palm Beach that nobody could talk, and anyone whose mobile phone rang would be kicked out, and – in Volusia – they controlled exactly what colour pens the counters were allowed (red) and where they could put their purses (next to the guard). It was no good: any count – even one carried out by machines – was going to be an interpretation. At one stage, sheriff’s deputies in Broward County took custody of 78 minute chads as evidence of potential ballot tampering. The police put them into an envelope marked: ‘Crime. Found Property’. An official was even accused of eating the stuff. That’s what happens when numbers suddenly reveal themselves to be the relative props they really are.
Even so, we have yet to see much resistance to the number-crunchers in the UK, except for the decision to stop publishing school league tables by the new assemblies in Northern Ireland and Wales, but it’s a different story in France – especially among economics students. ‘It was in the beginning a modest initiative, almost confidential,’ wrote the prestigious French newspaper Le Monde in September 2000. ‘It has now become a subject of an important debate which has created a state of effervescence in the community of economists. Should not the teaching of economics in universities by rethought?’
What had happened was that a small group of students had put a petition on the Web protesting against the ‘uncontrolled use of mathematics’ in economics. They claimed that the result was that mathematics had ‘become an end in itself’, turning economics into what they called an ‘autistic science’ – dominated by abstractions that bore no relation to the world as it really was.
Within two weeks, the petition calling for reconnection with the real world had 150 signatures, many from students at France’s most important universities. Soon newspapers and TV stations all over France had picked up the story and even senior professors were starting a similar petition of their own. By the autumn, the campaign had led to a major debate at the Sorbonne and the French education minister, Jack Lang, had promised to set up a commission to investigate the situation, and come up with some proposals to change economics teaching.
Even if they haven’t managed such a spectacular turnaround on this side of the English Channel, the UK press had been more vigilant than usual about such matters. The Independent on Sunday revealed that Department of Health figures for casualty waiting times don’t start when you arrive at hospital, but only measure the gap between when patients are first seen and when they are actually treated. And so it was that one 88-year-old who waited 24 hours before being admitted to hospital was officially recorded as waiting only 30 minutes.
Then there was the story of the government target to cut the time patients are allowed to wait on trolleys to four hours. Hospitals turned out to have circumvented the problem by buying more expensive trolleys and redefining them as ‘mobile beds’. Or the tale of the Kent police who promise to answer the phone within two minutes, but actually just do so by switching you through to a call centre that keeps you hanging on the line for ages.
The slow suspicion is dawning on us poor counted and measured people that this kind of stuff doesn’t work. The big lies of numerical targets, that they measure something real and that they can be manipulated to create change, are becoming ever more apparent. What isn’t yet clear is what we can do instead, how we can retain some of the human complexity that measuring tries to remove, how we can create institutions that build real human contact rather than the fake numerical kind.
Having lived with this problem for some years now – not to mention being grilled on radio about it, often thrown into the lion’s den with statisticians and census managers – I’ve a feeling it’s all about telling stories and asking difficult questions. Telling stories, because they can often communicate complex, paradoxical truths better than figures. Asking questions because they can devastate most political statistics. Yes, the carbon monoxide rate has reduced, but is the air cleaner? Yes, our local university professors have produced a record number of learned published papers, but is their teaching any good? Yes, the exam passes top the league tables, but what about the education? Are the children happy? Can they deal with life? Numbers and measurements are as vulnerable as the Emperor’s New Clothes to the incisive, intuitive human question.
The closer any of us get to measuring what’s really important, the more it escapes us, yet we can recognize it – sometimes in an instant. Relying on that instant a bit more, and our ability to realize it, is probably the best hope for us all.
Bizarre units of measurement No. 13
Dram
(Ancient unit of measurement for drink. 1 fluid dram = one hundred and twenty-eighth of a pint in the US, or one hundred and sixtieth of a pint in the UK.)
* * *
Number of Americans injured by supermarket trolleys every hour: 5
Average duration of sex in the UK: 21 minutes (compares to 14 minutes in Italy and 30 minutes in Brazil)
British school children who think Adolf Hitler was British prime minister during the Second World War: 4 per cent
If you enjoyed The Tyranny of Numbers, why not try …
Buy the ebook here
Buy the ebook here
Index
The pagination of this electronic edition does not match the edition from which it was created. To locate a specific passage, please use the search feature of your e-book reader’s search tools.
abacus 10, 11
Abbott, Charles 64
accountancy 11–13, 38–9, 49, 125, 132–3, 135, 143–5, 217
accuracy 7, 51
Adams, John 201, 208, 212
Addington, Henry 24
air pollution 46, 182, 192
Analytic Hierarchy Process 218
Anderton, James 52, 53
animals 1–4
Annals of Agriculture 27
anthropology 41, 54
Antidote 101–2, 106
Anyon, James 38, 47
Aquinas, St Thomas 14, 170
Arabic numbers 7, 10
Arnott, Dr Neill 66, 72
arts 96, 104, 106, 155–6, 176
Ashley, Lord 52, 69, 75, 76–7, 80
Atlantic Monthly 180
auditing xv, 39–41, 47, 49, 50, 217
ethical 140, 142–3
social 143–8, 150–1
Augustine of Hippo, St. 5, 14
autism xiii, 221
Average Man 44–5
Babbitt, Bruce 48
Bacon, Roger 8
Badiou, Alain 14
Barbier, Edward 204
Bazalgette, Joseph 78, 80
Beddoe, John 108–9
Ben & Jerry’s 141, 143, 145–6
Bentham, Jeremy xv, xviii, 16–36, 64, 66–8, 82, 125, 155, 216, 219
Bentley, Tom 106
Beveridge, Sir William 162
Bible 8
Blackstone, William 23
Bloomberg, Michael 149
Bloomsbury set 154, 155, 159
Blueprint for a Green Planet (Pearce) 204–7
Bly, Robert 91, 177
Board of Health 75–6, 78–9, 112
Board of Statistical Research 115
Body Shop 143, 145, 147
Booth, Charles 108, 110–26, 156, 184
Booth, William 110, 121
Bottomley, Virginia 89
Bowring, John 19, 29
Boyson, Dr Rhodes 51–2
Brandon, Nathaniel 99
Bretton Woods 168
Brooks, David 95
Buchan, James 11
Buchanan, Colin 201, 202
Burke, Edmund 18
business 106, 128–52, 217–18
corporate culture 103, 148
and emotional literacy 103–4
ethics xviii, 136–43, 147–8, 224
and information 128–36
intangible assets 58–9, 148, 222–3
management xvii, 129–32, 140, 151, 220
performance measures xvii, 140–1, 148–51
profits xii, 11–12, 59, 129–30, 140–1
reengineering 131
social auditing 143–8, 150–1
targets 49–50
Butler, R.A. 175, 181
Buxton, Jedediah xiii, xix, 221
calendar 7, 61
Canfield, Jack 95, 99
Cannadine, David 216
Carlyle, Thomas 18
Carroll, Lewis 51
Castle, Barbara 200