A vision of the future - in the official figures

The head of the Government statistical service is a staunch defender of its integrity, writes Diane Coyle

Tim Holt, head of the Government's statistical service since July, has inherited an unfinished agenda. It is an ambitious one: to rebuild public confidence in official statistics.

According to his predecessor, the ebullient Australian Bill Maclennan, the credibility of official statistics has been seriously undermined by the public perception that the unemployment figures were distorted by political pressures. Dozens of changes in the rules for claiming unemployment benefit have steadily pushed the total downwards.

Mr Maclennan insisted, before he quit the job to return to Australia, that the UK must switch to an internationally-accepted measure of unemployment - based on a monthly survey. His successor has brought this change close to fruition.

Two weeks ago Dr Holt announced the results of a Central Statistical Office study of the costs and benefits of alternative techniques for collecting the superior survey-based figures.

The statisticians' preferred method would cost an extra pounds 7-8m on top of the pounds 5-6m already spent on a quarterly survey.

After a period of consultation the CSO will make its recommendations to the Chancellor of the Exchequer. The decision about whether this country will have credible monthly unemployment figures to supplement the quarterly ones will rest with Kenneth Clarke.

Dr Holt is careful to echo the ministerial view that any change would have to give value for money. ''Without any cost constraint people always want more information,'' he said, adding: ''It is taxpayers' money. We need to be sure it is spent wisely.''

However, his own opinion is pretty clear: ''Direct government expenditure on labour market issues is pounds 12-13bn. You have to put the cost of a monthly survey in that context. We have a responsibility to put in place the information for those billions to be managed properly.''

Dr Holt insists that there is no undue political interference with official statistics. Even so, he says, public confidence is something that always has to be nurtured.

He gives every sign of being a staunch defender of the integrity of the Government Statistical Service he now heads. ''Reliable statistics are a major cornerstone of democracy,'' he stated.

Mr Maclennan cast a shadow over his own reputation as a beacon of integrity by suppressing just over a year ago an article in Social Trends, one of the CSO's flagship annual publications, by Muriel Nissel, one of its former editors.

Mrs Nissel's article criticised political influence on official statistics during the 1980s. She said Mrs Thatcher's Governments had cut spending - especially on politically sensitive social statistics - so much that their quality suffered.

Writing in the Independent, she said: ''Unless the Government is prepared to support a statistical service which publishes uncomfortable as well as comfortable facts, a democratic society will not have confidence in it.''

Dr Holt said: ''It has been government priority for the past seven years to strengthen the statistical information available for the management of the economy.''

Yet managing draconian cuts in running costs at the same time as the CSO's forthcoming merger with the Office of Population Censuses and Surveys will be one of the first tests of his skills as a defender of our statistics.

Like other government departments, the new Office of National Statistics will have to slash running costs by up to one-fifth.

The merger itself will allow some savings, and Dr Holt argues that using new technology in gathering and processing statistics is an important source of efficiency gains over time. ''We can maintain the level of quality and output,'' he said.

He does not see selling statistics as a big source of future income. The ONS will not turn into a commercial organisation, although the existing process of publishing some data through commercial partnerships will continue. He also mentions the possibility of charging for special analyses.

''Our fundamental priority is to get the statistics used. We will not be exploiting our monopoly position in holding the data,'' Dr Holt says. Statistics are a public good, he argues.

His more immediate priority is to make the 1 April merger a success. The management task is immense. Dr Holt, a former professor of statistics and Vice-Chancellor of Southampton University, has to cope with combining two big government departments without making either feel aggrieved, creating a new identity and achieving public recognition for the new organisation, moving to a new building and installing new systems.

In the longer run, he has a vision: he sees bringing together the most important economic and social statistics for the first time as an unparallelled opportunity. ''It will be the first opportunity in this country to paint a complete and coherent picture,'' he said.

A researcher wanting to know everything about, say, single parents or the elderly - from their benefits and incomes to their health and housing conditions - will find the information easily under one roof.

A business planning an investment will find regional figures on everything from wages and unemployment to demographic patterns.

The figures are already there, but hard to access - the separate pieces of the jigsaw scattered around several government departments.

Even more ambitiously, Dr Holt said, new information technology could eventually be a picture painted by the users of the statistics themselves, not an official identikit: ''Historically, we have collected, designed and published our product. In future there will be much more access to databases that allow users to design the statistical output they want.''

This democratic vision, distant as it is, would be the ultimate guarantee of freedom from political interference in the uses and abuses of statistics. Whether the resources and political will to realise Dr Holt's vision will be there is another question.