GSS Logo

High quality statistics, analysis and advice to help Britain make better decisions.
 

How to avoid common breaches of the Code of Practice

The most commonly occurring breaches of the Code of Practice for Official Statistics are pre-release statistics being sent out early, pre-release statistics being supplied to somebody who is not on the approved pre-release access list and statistics not being published at the required time of 9.30am. For each of these common scenarios, considerations and suggested corrective actions are set out below.


Pre-release statistics have been sent to people who should not have access to them

Considerations:

  • How many people have received the statistics in error and who are they?
  • Are the statistics high profile and / or market sensitive?
  • For how long have the recipients had access to the statistics before the error was discovered?
  • Have the recipients shared or discussed the statistics with others?
  • Can the offending email or statistics be recalled or deleted?
  • Was the correct security marking applied to the pre-release access email?  Could stronger wording or other signposting such as watermarking be used to make it clearer that the recipient is handling pre-release access statistics?

Corrective actions:

  1. Recall the statistics.
  2. If the statistics have been forwarded by somebody that was eligible to receive pre-release access, consider removing their pre-release access.
  3. Remind staff about the correct pre-release protocols.
  4. Strengthen the wording of all text accompanying pre-release material.
  5. Consider further training to educate staff about their obligations under the Code of Practice and to ensure that all those with pre-release access are properly briefed.
  6. Consider whether increased management control of the processes is appropriate, and action accordingly.

Pre-release statistics have been sent out too early

Considerations:

  • How many people received the statistics early?
  • Was the person that sent the statistics out working for the organisation that produced the statistics, or were they from another organisation?

If there are only a small number of trustworthy recipients, consider recalling the statistics and asking recipients to delete the email and confirm that they have done so.

If a large number of recipients are involved, consider whether publication can be brought forward to give the maximum allowed 24 hour pre-release access. Change the publication date on your organisation’s website and the Release Calendar on GOV.UK as soon as possible

Corrective actions:

  1. Recall the statistics.
  2. Bring forward the release date of the publication.
  3. Remind staff of the requirements of the Code and correct process for sending out pre-release information.
  4. Review the processes involved in sending out pre release information, and make any changes that are needed.

Statistics have been released before their scheduled publication date

Considerations:

  • How sensitive are the statistics and how long is it before the scheduled publication date?
  • How many people are likely to have accessed the statistics?
  • Has pre-release access to the statistics been restricted? Should you ask people with pre-release access not to disclose or discuss the statistics until further notice?

Corrective actions:

  1. Withdraw the statistics as soon as possible.
  2. Bring forward the time of the general release.
  3. Issue a statement on your organisation’s website alerting users to the problem.

Statistics have been released after 9:30am on the day of publication

Considerations:

  • How sensitive are the statistics and how long is the delay likely to be?
  • Has pre-release access to the statistics been restricted? Should you ask people with pre-release access not to disclose or discuss the statistics until further notice?
  • Can social media channels be used to acknowledge or apologise for the delay?

Corrective actions:

  1. Consider emailing key users a copy of the release.
  2. Issue a statement on your organisation’s website alerting users to the problem with a proposed resolution time if this is known.
  3. Consider whether there is another way to publish the release.