The outbreak of the War of Numbers can be traced back to the early nineteenth century. Before that, there was a benign discipline, quaintly christened ‘political arithmetic’, which offered some scant numerical evidence about population, life expectancy and finance. But the novel idea that the state should routinely collect and publish ‘statistics’ blossomed in the Victorian spring and sparked a furious and protracted struggle for domination over data that continues to this day.
One of the forefathers of statistical science, Sir Francis Galton, regarded statistics as beautiful but combustible, warning that they should not be brutalised but delicately handled and warily interpreted. Benjamin Disraeli, however, was less enamoured. In 1847 he wrote of how delightful it would be to suppress their use and sack the imbecile who ran the Board of Trade’s statistical department. It is doubtful that Disraeli ever did use the phrase ‘lies, damned lies and statistics’, but what is certain is that there were those who wished to smother the infant science in its cradle. Nevertheless, guided by the principles of the Enlightenment and inspired by the belief that the solution to society’s problems lay in the hands of men, reformers saw statistics as the instrument for creating a better world. Britain witnessed a fevered, almost obsessive period of counting and categorising (see ‘J is for Justice’).
The Statistical Society of London, founded in 1834 and later becoming the Royal Statistical Society, had as its first emblem a wheat sheaf, representing a bundle of facts, bound by a ribbon with the motto ‘to be threshed by others’. Members were told to confine their attention rigorously to facts stated numerically and arranged in tables, carefully excluding all opinions. But it quickly became obvious that total immersion in the pure waters of numbers was an untenable position. The motto and the prohibition were dropped in 1857.
From this flowed a rather unseemly free-for-all in Whitehall, with statistical squabbles over methodology, accuracy and interpretation spreading between departments and across generations. It took a real war and a fat cigar to instil some order onto this bureaucratic chaos. In 1941, Sir Winston Churchill demanded a Central Statistical Office ‘to consolidate and make sure that agreed figures only are used’. Consolidation and agreement, however, would remain in short supply as the various departmental statistical branches competed to produce numerical justification for the experimental ideas that would shape post-war Britain.
Often their data proved to be unhelpful in justifying government’s policy ambitions. The ‘facts’ had an irritating habit of getting in the way of the ‘good stories’ ministers wanted to tell. It was Churchill again who famously told a prospective parliamentary candidate: ‘When I call for statistics about the rate of infant mortality, what I want is proof that fewer babies died when I was Prime Minister than when anyone else was Prime Minister. That is a political statistic.’
Harold Wilson had respect for data and those who assembled it. He had spent the war working as a statistician for the coal industry but, as Prime Minister, he too recognised the tension between statistics and politics. In a speech to the Royal Statistical Society in 1973 he said he had learned that, ‘while you can always get someone to find the answers to the questions, what you need in government is the man who knows the questions to the answers.’
The balance of power between elected ministers and qualified statisticians, between politicians and scientists, shifted palpably with the arrival of Margaret Thatcher. She wanted all her numbers to be political and invited an ideological soulmate, Sir Derek Rayner, the man from Marks & Spencer, to conduct a value-for- money review of the Central Statistical Office. He helpfully concluded that it was ‘too heavily committed to serving the public at large’. In Sir Derek’s view, ‘information should not be collected primarily for publication [but] primarily because government needs it for its own business.’ Overnight, troublesome number crunchers were stripped of their public responsibilities. Many lost their jobs as the government wiped out a quarter of its statistical service. Instead of providing ammo for voters, Whitehall stats departments would do what they were told by ministers. If this was a key battle in the War of Numbers, it was one which saw the boffins crushed.
At 4.30 on the afternoon of Wednesday, 10 June 1981, more than 200 statisticians gathered at the London School of Hygiene and Tropical Medicine to lick their wounds and vent their frustrations. The discussion was described as forceful, the atmosphere electric. The collective noun for statisticians is ‘a variance’, but on this issue they were united and determined. The long fight back would be conducted neither in Whitehall nor Westminster. The statisticians would go global.
The collapse of the Soviet empire in the early 1990s had created newly independent states which, it was hoped, would abandon the corrupt centralised systems of the Communist era, including the control and manipulation of official numbers. Here was an opportunity for the statistical profession to regroup on the moral high ground. Under the auspices of the United Nations in Geneva, plans were set in motion for renegotiating the balance of power.
In 1994, the UN published what were called the
The UN statement was a direct challenge to governments. Statistics were too important to be left open to the potential abuses of elected representatives. A year later, at the headquarters of the Royal Statistical Society in north London, the shadow Home Secretary Jack Straw admitted that political control of numbers had deeply damaged the relationship between the governed and the governing class. ‘There can come a point where the cynicism goes so deep that it corrodes the foundations of our political system, leading to a wholesale lack of confidence in the system,’ he warned. ‘I believe that we are dangerously close to that position today.’
The Labour Party, frustrated particularly by the way unemployment figures had been ‘massaged’, was promising a new independence for those compiling government data. Statistics had become ‘the hard and brittle currency of politics,’ Straw conceded. Controlling the numbers had become key to the partisan battle for the hearts and minds of the electorate.
When Tony Blair and Gordon Brown moved into Downing Street in 1997, their luggage contained a promise of evidence-based policy. The idea that cold facts rather than hot passions should lead government was further encouragement for statisticians who looked forward to seeing their work and their influence increase. A year after entering office, the new Labour government published
A decade later, in 2008, the rather toothless watchdog created ostensibly to restore faith in government data, The Statistics Commission, had to admit that the use of official numbers ‘continued to be driven largely by departmental requirements’. It also noted that government was able to spin the statistics because it often controlled both the release of the data and the ministerial reaction to it.
I witnessed the game being played at the Home Office when the crime figures were published. Journalists would be invited to a ‘lock-in’ at the department, quite literally. The press would be ushered into a windowless room and instructed to turn off all telephones. The doors would be shut as a news release was thrust into our hands, proclaiming the good news of falling crime and delighted ministers. Then the assembled hacks would thumb through the ‘official’ stats, searching for what they regarded as the real story. Buried in the pages of numbers and charts would be some narrow category of crime which was still increasing, evidence that equated with voters’ settled view: crime was out of control and politicians were deceitful toads. Upon being freed from our enforced purdah, that is the story we would tell. The more ministers struggled to convince the public that overall crime was falling, the less the electorate trusted them. In seeking to spin the numbers, they destroyed the reputation of government statistics and themselves. What was particularly galling for both politicians and statisticians in the Home Office was that the figures really did show that crime was falling. Quite a lot. And in almost every category.
However, the real damage to government credibility was being felt at the Treasury, where the Chancellor Gordon Brown was increasingly concerned that trust in economic statistics was denting international confidence. In a brief statement in Parliament towards the end of 2005, Mr Brown announced he would make ‘the governance and publication of official statistics the responsibility of a wholly separate body at arm’s length from government and
