Britain’s Labour government: statistics, damned statistics, lies

A research paper prepared by leading academics accuses the Labour government of manipulating statistical data in order to mislead opinion on its achievements in government.

A good place to bury bad news? Hiding the detail in the geography on The Labour Party’s website”, was produced by a team of geographers led by Professors Danny Dorling at the University of Leeds and Ron Johnston at the University of Bristol.

The paper begins by drawing a comparison with the fictional Winston Smith in George Orwell’s 1984. Employed in the Ministry of Truth, Smith “spent his working days re-writing history in ‘Newspeak’ so that it favoured ‘the Party’.” As part of his job Smith would systematically readjust the Ministry of Plenty’s figures, so as to present them in a more favourable light.

“What was the stuff of fiction in 1949 is commonplace ‘spin’ in 2002” the research paper’s authors write. The anonymous persons responsible for maintaining the Labour Party’s web site routinely massage statistical data in order to present the government in the best possible light. The team’s findings are based on an appraisal of statistical information featured on “What Labour’s done in your constituency” from the Labour Party web site, (www.labourparty.org.)

Set up prior to the 2001 general election, the site contains a list of statistical data for each of the 641 parliamentary constituencies in England, Scotland and Wales. A voter is meant to be able to type in the postcode of their constituency and receive an up-to-date profile of the benefits brought to their particular area by a Labour government.

The figures, which the party claims are drawn from publicly available data, are all meant to show that things have “only got better,” as the government had pledged.

The paper shows that Labour’s choice of areas of improvement are highly selective. Indices chosen include reductions in crime, expansion of staffing in hospitals and schools, decreased mortgage repayments and falling unemployment. No information is presented on Labour’s much publicised pledge to cut levels of child poverty. But this is only the start of Labour’s stacking the decks. Whilst there is no doubt, for example, that unemployment has fallen and mortgage repayments are at their lowest for years, the authors found that in instances where improvements could not be shown using the same criteria, the source of the data used was changed.

“If an indicator had not improved for one time scale then the time scale was changed for that constituency to one during which conditions had improved,” the report notes. Different spatial indicators, using a broader aggregate, were also used in other instances. The end result was to paint a picture of universal improvement in every area cited, irrespective of where a person lived.

On unemployment figures, local constituency figures were used in each case, except where unemployment has actually risen, where figures for the sub-region were substituted.

In the case of decreased hospital waiting lists, two alternative time periods were used (1997 or 1998) depending on which provided the better picture. Reported increases in nursing staff were also based on regional figures, rather than the smaller Health Authority statistics, helping mask those particular districts where improvements had not been made.

The authors point out that whilst none of the information published on the Labour party site is “untrue in the strict sense,” the way that it is put together is “disingenuous” overall.

This raises the danger that such information is being disseminated through the media, local Labour Party branches and other means as reported “fact.” The authors ask where such practices end? What will the Labour Party do when unemployment begins to rise, or mortgage rates start to go up, as they surely must eventually? Just how far is Labour prepared to go in creating a “social statistical utopia”?

The research paper also examines the figures from the standpoint of election strategy. It divides the country into three distinct political entities—Conservative and others, Old Labour and New Labour in order to establish whether the statistics cited are aimed at improving Labour’s political fortunes in key marginal areas. This would mean better statistics being reported in New Labour areas—marginal constituencies won from the Tories in 1997, etc.—as Conservative and Old Labour seats would be considered either unwinnable or unloseable respectively.

The resulting date shows no such pattern, although the choice of indices focussed on would tend to be those most of concern in New Labour areas, the authors note.

The report explains, “Few of the indicators provide clear evidence of change between 1997 and 2001, however: most provide static pictures only.” Just seven indicators showed any change since Labour came to power, it continues. Five of these were in those marked old labour, “and in all five the data were manipulated, with different spatial and/or temporal scales being deployed to present the changes in the best light.”

That is not to say there is no trend. The authors note that, without national comparisons, a person living in an Old Labour area (in working class areas of major towns and cities) would not be “able to discern that their party’s actions over four years in government had been much more to the advantage of those living in areas which in the recent past had not supported Labour (i.e. the more middle class areas of the country)—and thus to ask whether ‘their’ party had really been serving ‘their’ best interests!”

The paper hints at the political considerations behind such statistical manipulation. It concludes that “Rather than appearing to be a necessary series of occasional white lies, it is beginning to look as if the provision of this distorted picture of the local geography of Britain is a longer-term party strategy.”

In placing Labour’s statistical information under the microscope, the paper’s authors have provided an invaluable service in righting the government’s campaign of misinformation. They themselves note that Labour’s web site assumes greater importance as a propaganda tool given the decline in party membership. However the party’s massaging of figures is put down to its desire to “be all things to all men and women.” It would be more correct to say it flows from the government’s need to obscure the impact of its pro-big business policies on the working class and how this fosters social inequality.