Access to statistics before they are released should be sharply curtailed, the UK Statistics Authority has recommended.

It proposes ministers and officials should get just three hours pre-release access, and journalists none at all. For an hour after the statistics are released, the only commentary available should come from statisticians, not ministers or press officers.
 
The recommendations are designed to increase public confidence in statistics and give ministers or advisers less time to cook up a positive spin and influence how the statistics are covered by the press. But the authority acknowledges that under the law it does not have the power to enforce these changes and, short of new legislation, suggests instead that ministers should commit themselves to following the recommendations.
 
As things stand, variable numbers of people within Government departments in England and Northern Ireland (an average of 21, with a range from 8 to 55) have pre-release access of 24 hours to statistics due to be released at 9.30 am the following day. In Scotland and Wales, the pre-release times are much longer, five days.
 
A large number of journalists also have pre-release access, on the grounds that it helps them to prepare more accurate stories. But the authority takes the view that this access is unjustified, and creates a “perception of collusion” between the press and Government press offices, and a situation in which journalists might know more than people they were interviewing.
 
It argues that journalists have no more right to an early sight of statistics than Parliament, which gets no such access. The only exception it envisages is when journalists might be given access in a “lock-in”, in which they see the statistics but cannot communicate with the outside world until they are officially published. This arrangement is already widely used for the publication of important reports.
 
In recommending that statisticians should be given the first chance to comment on statistics the authority is following US practice. The hope would be that getting the first word in would frame the succeeding discussion in a statistically sound way, leading to better understanding and avoiding the situation where officials pore over the figures to find a positive angle which then forms the basis of a press release.
 
The three-hour pre-release time represents a compromise. Sir Michael Scholar, the authority’s chairman, has made it clear that he would favour no pre-release access at all. But this is probably a bridge too far, so the authority has opted for the same three-hour period recommended by the Treasury Select Committee in a 2006 report.
 
Along with the changes in timing, the authority wants to see a change in culture. “There is currently an expectation, inside and outside Government, that ministers and departments must be able to give instant comments when new statistics are published” it says. “It is our view that this should change and more regard should be paid to a central principle of good statistical practice – equality of access. Current practice has developed as a consequence of pre-release access but it should not be used to defend it.”
 
If the three-hour limit is agreed, the timing of the release of statistics may need modification. Currently it is set at 9.30 am, but might be shifted later in the day to enable those still entitled to pre-release access to make use of it in normal office hours.
 
As for journalists, they will have to work faster, or delay their analyses for a day or two in the case of complex sets of statistics such as school examination results. “The public value of the school statistics will be no less a few days after the raw results become available; indeed, it would be helpful if the statisticians’own analysis of the results were to become the focus of media attention initially” the report says.
 
It further argues departments could help by putting the statistics in a form journalists can more easily use – the media’s wishes “should be seen as a major customer requirement by the departmental statisticians”. There is, of course, a problem with this: journalists always want to construct league tables, while statisticians do not. 
 
Straight Statistics supports the view that improving public confidence demands changes in pre-release access. The three-hour limit proposed is the compromise most likely to win political support.
 
Sir Michael Scholar said: “Equality of access is a very important principle of statistical good practice. All the parties to public debate should, as far as possible, have the same information available at the same time, and only the minimum number of people should see statistical results before they are published.
 
“The current 24-hours advance access by Ministers and their advisers contains too many dark hours during which no-one can see what is happening. In Scotland and Wales, the current arrangements allow five days advance access for devolved statistics, which is inconsistent with international best practice. 
 
"We believe that the recommendations in our report will minimise the opportunities for political influence or exploitation, and will help to build public confidence in the independence of the statistical system. 
 
The president of the Royal Statistical Society, Professor David Hand, supported the recommendations. He said: “Public confidence in official statistics is at appallingly low levels. The most recent official survey carried out by the Office for National Statistics showed that only one in five people thought that official figures were compiled without political interference. 
 
“Extensive ministerial access to statistics before publication can only fuel the public lack of confidence. Severely curtailing pre-release access to three hours throughout the UK, as the Statistics Authority recommends, would send a strong and clear message that public confidence is paramount. The Society itself has argued for no pre-release access as is the case in a number of countries.”