« Previous | Contents | Next »
Listen
Pre-release access arrangements
59. With few leaks or trailing of statistics due to pre-release access, the rules around this appear to work well in Scotland. In light of this there has been very little public debate in Scotland around early access by Ministers and policy officials to statistical data in its final form.
60. Some UK professional statistical bodies have commented, through the HM Treasury consultation, that pre-release access contributes to the perception of Ministerial interference in statistical outputs. Notably, the Royal Statistical Society has recommended that there should be no pre-release access given to Ministers and policy officials.
61. The relevant international principle here is that access should be equal 8. That might appear to rule out any pre-release access. However, in practice pre-release access occurs in many countries, though where it does the length of time granted, and conditions under which early access is granted, varies considerably. The European Union Code of Practice specifically recognises that pre-release access occurs and it requires that it should be limited, controlled and publicised - the existing UK Code states much the same.
62. The Statistics and Registration Service Bill makes the provision for secondary legislation to set out pre-release access arrangements for statistics in their final form. Scottish Ministers can decide on the arrangements for Scottish devolved statistics.
63. Although it is not envisaged that this secondary legislation will include details about access to administrative and management data, which Ministers and senior officials require as part of the daily business of running government departments, or over early access to statistics for quality assurance purposes, we want to consider all types of data here and explore how they should all be managed.
64. We feel that pre-release access has several legitimate benefits and there should be less concern about arbitrary time limits and more about what happens in the pre-release access period.
65. The case for early access to statistics produced in government is strong as:
- A significant part of the pre-release period is used to ensure that policy officials properly quality assure the statistics and, moreover, that they and Ministers understand the statistics;
- Ministers are expected to comment on an informed basis at the time that statistics are released. They may also be called on to defend the accuracy and reliability of the statistics;
- Furthermore it cannot be in the interests of good government for Ministers to have had insufficient time to consider the policy implications of statistics;
- Pre-release access is important from a statistical integrity point of view. Ministers can react to statistics whether they have seen them in advance or not and may often be called on or wish to comment on the policy issue at any time;
- Many of our statistics are derived from data that may actually constitute management information. These are government data to which public access is being allowed - in a very positive and pro-active way. But it would be both futile and wrong to say then that government access should be limited prior to public release.
66. The arguments against pre-release are:
- Lengthy pre-release access increases the risk of, and perception of, political interference - meaning attempts to influence the timing, emphasis given to specific aspects of content and interpretation;
- It may be perceived that Ministers influence the context (rather than content) thereby reducing the impact of politically sensitive results; and
- It may also be perceived that pre-release access gives Government an unfair advantage in presentation, increasing the risk of leaks and breaches, leading to selective early release of information to encourage favourable media coverage.
67. The case for pre-release access is convincing but the arguments against are also valid. Therefore, we conclude that pre-release access should be retained but that some restriction is desirable, to provide assurances that there will be no trailing of statistics and selective early release; this would fully meet international standards. However, arbitrary time limits do not in themselves prevent misuse and the main effect of these is simply to inconvenience Ministers and policy officials.
68. The Scottish Administration is open about its current release practices and details about standard pre-release access arrangements are outlined in the Scottish Executive's UK Code of Practice compliance statement 8. Exceptional pre-release access details were formerly published on the Scottish Executive website, but seemed to attract little interest so the specifics were removed without public reaction.
69. The current arrangements for pre-release access do not make enough distinction between the different types of statistics used in government.
70. There are potentially many different types of statistics, which should be managed differently but we need to avoid too many separate arrangements. We consider the following to be the main distinctions, although we would particularly welcome comments from users on this proposal:
- Market sensitive statistics (as at present) where release is rigorously controlled;
- Statistics that have been produced from management information or administrative data;
- Survey statistics collected from businesses and the public;
- Secondary release statistics, these are detailed statistics produced from data already published;
- Statistics for Scotland sourced from UK wide surveys;
- Background data and other detailed information including for example Neighbourhood Statistics.
71. Pre-release access arrangements for all these types of statistics will not be detailed in the secondary legislation but we want to ensure that our own guidance on all of these is clear.
72. We propose to:
- Draw up detailed plans for pre-release access that will reflect the different types of government statistics.
Questions
What do you consider to be the different types of government statistics?
Should different pre-release access rules apply to different types of government statistics?
« Previous | Contents | Next »