Why is the government trying to restore faith in statistics?
The public is increasingly questioning why there seems so often to be a gap between their own experience and the statistics reeled off in government reports. Take crime, for instance. According to the British Crime Survey, violent crime has fallen to its lowest level since the survey began. But the dramatic 43% drop in violent crime since the 1995 peak recently reported in the survey – which questions 40,000 adults about their experience of crime – does not take account of serious crimes that are too “small statistically to measure” (such as murder and rape), or crimes against people under 16. And in fact, crimes reported to the police suggest a different story: the most recent figures, for the final quarter of 2004, show a 10% rise in crimes of personal violence on the same period in 2003. Of course different measures will give different results, and governments always bend statistics to fit their aims. But New Labour’s reputation for spin has increased general scepticism. “Not since Robert Maxwell told us that his business was solvent have numbers been so effectively bent to fit the story,” says Jeff Randall in The Daily Telegraph.
What about economic statistics?
Inflation has been another source of debate. In a time of soaring fuel and council-tax bills, many people feel that inflation is far higher than the officially reported consumer price index rate, which currently stands at 1.9%. While a single inflation measure couldn’t possibly reflect the personal inflation rate of every individual in the UK, statisticians do use various techniques that have the effect of keeping the measure on the low side. For example, there’s ‘hedonics’, where if a product (ie, a computer) improves in quality, then it’s judged as having fallen in price, even if it still costs the same amount, because you are seen to be getting more for your money. Then there’s ‘substitution’, where, for example, a rise in the price of chicken is ignored because consumers are assumed to have switched to a cheaper meat, such as beef. There are similar problems with jobs data. Officially, unemployment in Britain may stand at 5.4%, one of the lowest rates in Europe. But if you take account of the UK’s “inactivity rate” – those of working age who are signed off sick or listed as ‘economically inactive’ – that number is in fact closer to 21.2% of the working-age population.
Why does this matter?
Statistics are important because they help provide a picture of what’s happening in the country. If the results are skewed to give people an image of what politicians wish was happening, rather than what is actually going on, policies will be founded on inaccurate information. This is particularly important in the economic sphere. And it’s not just disgruntled members of the public who are worried about the quality of statistics. In 2005, the Bank of England took to producing its own estimate of GDP growth after a series of gaffes and revisions to National Statistics figures, and suggestions of political interference hit confidence in the official economic growth figures.
Is the new watchdog likely to improve matters?
With the appointment of an independent statistical watchdog in Sir Michael Scholar, the Government argues that it is backing up its promise to make the process of publishing national statistics a more transparent one. But the move is unlikely to stop this or future governments from bending the truth about the state of the country – the most contentious statistics, such as crime, immigration and school’s data – will still be outside the control of the Statistics Board, notes Philip Johnston in The Daily Telegraph. The government will still also be able to hang on to its privilege of reviewing statistics well before they are passed onto the public. In the US, the government is afforded just a few hours between when it reviews national statistics and when they are presented to the public. This is set to be cut from 42 hours to 24 hours here in the UK, but, notes Johnston, that leaves plenty of opportunity to “get your ducks in a row”. All of this means that the most important thing to look at when you are interpreting statistics is not the data itself – but still who it comes from and what their agenda is.
It’s not just the UK
Taking account of the real rate of GDP growth over the course of the last 20 years reveals that official reports have consistently painted the US economy as far more healthy than it really is. The economy was meant to have been in bullish form for most of the 1980s and 1990s, but besides the recession in 1990/1991, troubling downturns in 1986 and 1995 were missed by the formal GDP reports, which have been made unreliable by upward biases the government has been building into its forecasts since the 1980s, says John Williams on Shadowstats.com. We might also reconsider Alan Greenspan’s reputation for saving the economy’s skin after the market dipped during the tech burst – the 2001 recession was longer and deeper than the government reported, dragging on into 2003. According to Mitchell, the US economy has already been in recession for some time, suffering a negative growth rate of just over 2%. And when he measured inflation using the old CPI method used 20 years ago, he found US consumers are currently enduring a 10% growth in inflation, rather than the 1.9% reported.