How do the three major political parties plan to improve public trust in official statistics? A meeting organised by the Royal Statistical Society today suggests that they haven’t yet given the issue the highest priority.

Central Hall Westminster (capacity 380) was almost full to hear the views of Francis Maude (Conservative), James Plaskitt (Labour) and Vince Cable (Liberal Democrats). But audience interest wasn’t really matched by the degree of political engagement.
 
Mr Cable arrived late, Mr Maude left early, and Mr Plaskitt, who earns full marks for attending throughout, isn’t any longer a minister. He is a member of the Treasury Select Committee and was doing the gig as a favour to the Cabinet Office. None had a party platform to present.
 
So what emerged? Mr Plaskitt believed that things had improved after the 2007 Act setting up the UK Statistics Authority, but he questioned the role of the authority acting as both “scorer and umpire” of statistics in the UK. He suspected further legislation might be needed. That would include putting the Code of Practice on a statutory footing.
 
He didn’t favour abolishing pre-release access to statistics. (None of the three did – no real surprise there.) Actually, he believed that extending it by including the chairs of all parliamentary committees in the pre-release access list would improve matters by providing a “belt and braces” reassurance that ministers could not abuse their position without being detected.
 
He also favoured a system of “kite-marking” so that every official and national statistic published would include a certificate of quality. (You couldn’t actually call it a kite-mark, as the British Standards Instititute defends that trade-mark with ferocious zeal.) He supported the 2011 Census, and argued that its benefits would exceed its costs.
 
Mr Maude disagreed. He said the cost, almost £500 million, would be excessive, the reliability of the data poor, the questions intrusive and filling in the return burdensome. It was 32 pages long, compared to 20 pages in 2001, when 10 per cent of the data had been “inputed” rather than collected.
 
“Compliance was already poor and is going to suffer substantially, while the quality of the data is going to suffer from the burden and intrusiveness of the Census” he said. “There is a good case for scaling the whole thing back.”
 
He questioned the claims made that the forms really needed to go to the printers this March, a full year ahead of the census. “Is the printing industry under such stress that it needs 12 months to print the forms?” he asked. It seems a safe bet that if Mr Maude’s views represent Conservative policy, the Census is going to suffer a substantial shake-up should they win the election.
 
Mr Cable said the Census should be supported, but was puzzled why the 2001 version had produced wildly inaccurate figures for demographic data. “If we can’t make an accurate head count, something is obviously wrong” he said.
 
He declined to support “ring-fencing” the ONS budget. “If you get into the language of ring-fencing it won’t be the statisticians who come out on top, but departments with sex-appeal, like health” he said.
 
His final contribution was the most interesting. In response to a question about the difficulty of obtaining data when private companies are contracted to provide public services – such as PFI contacts, or Independent Sector Treatment Centres – he suggested the need for a Data Commissioner, along the lines of the Information Commissioner, to determine what data people were entitled to know.
 
“This is a major area of opaqueness” he said. “The answer lies in having a stringent test of proof that companies have to satisfy before they can withhold data.” The Data Commissioner would be the judge of when that condition was met.