Clause 10
Statistics and Registration Service Bill
3:00 pm
Theresa Villiers (Shadow Chief Secretary To the Treasury, Treasury; Chipping Barnet, Conservative)
Amendment No. 140 would turn the code of practice for national statistics into a code of practice that covered all official statistics. It is supplemented and supported by amendment No. 91, which imposes on those who produce official statistics a legal obligation to obey the code. Together, the amendments would abolish the two-tier system that distinguishes between national statistics and other official statistics and ensure that the full rigour of the reforms applies across all official statistics, whether they are produced by the ONS or by Departments.
Before turning to the substantive issues that surround amendments Nos. 91 and 140, to broaden the scope of the code in the way that I have described, I should like to take some time to deal with the technical issues that arise from this group of amendments. It is necessary to go through them to understand the restructuring of the Bill that amendments Nos. 91 and 140, and some later consequential amendments, involve.
The Opposition consider it to be a matter of concern that the Bill does not require anybody to comply with the code of practice. Amendment No. 91 would impose such an obligation. We feel that imposing a legal obligation would give real statutory force to the code and the new regime for statistics and strengthen the ability of the board to hold the producers of official statistics to account. It then must be decided on whom the obligation should be imposed.
The Bill envisages that the code of practice will apply to sets of numbers. In fact, the focus should be on the institutions that produce the statistics; there should be a code of ethics and good practice for producers. That point is strongly made by the Statistics Commission, which has emphasised the importance of the code’s applying to the activities of Government bodies, rather than providing a means of kite-marking individual sets of numbers. The code of practice should set out ethical guidelines for the way in which Departments operate in relation to statistical matters. The board should monitor the behaviour of the relevant authorities to determine whether they have complied with the code across the board, rather than focusing on certain sets of statistics.
Our amendments would give the board the supervisory and regulatory powers that, as the Financial Secretary acknowledged this morning, it lacks under the Bill. The Committee will note that paragraph (a) of amendment No. 91 would require the relevant authorities to consult the board on matters of the code’s interpretation. That would address a problem with the existing non-statutory code, which has been interpreted differently by the Statistics Commission and Departments. In its 2005-06 annual report, the commission said:
“Only a few parts of the present Code are of a kind that allow an independent judgment to be made about compliance. Much of the Code is non-prescriptive and requires no specific evidence.”
That ambiguity has led to differences of view about what the code means. The Treasury Committee heard from Mike Hughes, a director of the ONS, who commented on similar concerns expressed by David Rhind. Mr. Hughes said:
“I have great sympathy with David Rhind’s remarks about the code at the moment, because the code is an articulation of good practice rather than saying, ‘You should be doing this or that,’ so it is left to individual departments to decide how they choose to interpret it.”
The most important way to remedy that problem is, of course, to eliminate ambiguity, but the Opposition also feel that paragraph (a) of amendment No. 140, with its reference to the interpretation of the code, would be a visible signal of the importance of the board as guardian of the code and the institution with real authority to interpret what it means. That visible signal of the board’s authority would also be strengthened by an obligation to report breaches of the code to the board, and that is contained in paragraph (b) of amendment No. 140.
Turning to the last technical point before addressing the substantive question of scope, amendment No. 93 would require the board to report breaches of the code to Parliament. That is more of a probing amendment than the others. Throughout the debate, hon. Members on both sides of the Committee have emphasisedthe importance of parliamentary involvement in scrutinising the new structures. I am sure that the board will, in any event, refer to breaches in the regular reports that it is anticipated that it will submit to Parliament. However, it is important to have the opportunity to assess how the reporting process would work. We would have to interpret the reporting obligation on the board in a proportionate way, so that it would not necessarily have to write to Parliament every week, but this amendment emphasises the importance of parliamentary scrutiny of the board’s operation.
I shall turn to the main point at issue. We tabled our amendments because we believe that the board’s role should be consistent across statistics produced by the Office for National Statistics and Departments. The Bill gives the new board the obligation to oversee statistics that fall outside the scope of National Statistics, but it does not give sufficient power and authority for the board to live up to that responsibility for oversight. Ministers will retain the power to decide which statistics from their Departments should go forward for consideration in terms of becoming national statistics and which should not, leaving important Government figures out of the new framework. Effectively, the proposals would give Ministers the power to decide to what extent the legislation should apply to them.
There is no good reason why departmental figures, many of which are vital in assessing the performance of our public services, should be treated to a lesser regime. Much of the controversy about statistics over the past few years has focused on departmental practices and statistics, not on ONS ones. As the Royal Statistical Society has pointed out, there is nothing inherent in the social statistics that come from Departments that justifies a differential treatment by the new structures to be set up by the board. On the contrary, such statistics are a highly important part of political debate today.
Our position draws support from a range of sources. The Greater London authority data management and analysis group, for example, has stated:
“We reject the proposition that statistics which derive from management of public services are inherently different and cannot meet the requirements of the National Statistics Code of Practice. If statistics merit release for public use, especially if they are being used as performance indicators, then they merit inclusion as National Statistics.”
The Audit Commission has also expressed concern, stating:
“There is the risk of creating a two-tier system of statistics, which will be confusing to the general public and other stakeholders. Such a scenario would only seek to generate low public trust in statistics and reinforce the perception of political interference in the production and publication of statistics.”
Dr. Barry Leventhal, chairman of the Market Research Society’s census and geodemographics group agreed that all
“unnecessary distinctions should be removed between national statistics and other official statistics—such distinctions are unsatisfactory and confusing to users.”
Professor Denise Lievesley, chief executive of the health and social care information centre has stated:
“It is the IC’s belief that the code should apply to all official statistics.”
Sue Duncan, the Government’s own chief social researcher, has stated:
“It is important to recognise that much of the statistical system falls outside National Statistics and is therefore not subject to the same quality control, which has the potential to undermine confidence of the public (who won’t make the distinction between National and other statistics produced by departments).”
To put that remark in context, I should make it plain that she did not call for a total abandonment of the distinction between national and other official statistics.
Debrah Harding of the Market Research Society has said:
“The agreed definition of national statistics must be consistently and appropriately applied by government, the ONS, and the research and statistics community as a whole. Fundamentally the distinction must incorporate all official statistics including crucial policy areas such as health, education and crime which are currently produced outside of the ONS.”
Many have expressed concern about the role of Ministers in nominating statistics for the board’s consideration to become part of the National Statistics system. The British Urban and Regional Information Systems Association has stated:
“The implication is that there will be a list of statistical series that will be within scope and this will be decided bottom up by Ministers on a case by case basis. In effect Ministers are being given the job of deciding whether the legislation should apply to them. In our view, the Board should have responsibility for safeguarding the public interest.”
Jonathan Baume, general secretary of the First Division Association of senior civil servants, has said:
“The FDA believes that allowing Ministers alone to decide which statistics fall within the remit of National Statistics would undermine attempts to build a clear identity for independent statistics. It would either lead to confusion and add to lack of public confidence in the statistics published or it would create a two tier regime for government statistics which is undesirable and could potentially create difficulties for the government of the day.”
Professor Tim Holt has pointed out:
“The proposal that Ministers should decide what statistics are to be covered by the arrangements is unacceptable. We have seen the effect of this fragmentation of decision making to individual Ministers under the current arrangements and it is unsatisfactory. A legal framework will not command public confidence if Ministers can decide whether or not it applies to them and activities within their departments.”
The Opposition believe that Ministers are unlikely to be keen to nominate their departmental statistics to go into a new scrutiny system. According to Dr. Fellegi, the chief statistician of Canada, allowing Ministers to decide what should or should not be designated a national statistic
“would undermine the whole idea of statistical independence...surely the likelihood that, should they do so, their statistical activity would be subject to audits is not a very strong incentive to opt in.”
If this aspect of the proposal is not changed, Dr. Fellegi feels that the whole exercise amounts to no more than “tinkering”and would fail to solve the problem of public trust in statistics. The academic and statistician, Lord Moser, whose views we have heard a number of times in Committee, told the Treasury Committee that that aspect of the Government’s proposals was a basic flaw.
The RSS has stated that the extension of National Statistics over the past six years has been patchy. Many people, including the Committee chaired by my hon. Friend the Member for Sevenoaks, have expressed concerns about the number of statistics relating to areas such as crime, education, health, and social security that are outside the scope of National Statistics. The Statistics Commission has stated
“We believe it is essential for the new arrangements to relate to all official statistics, not just those which are currently the responsibility of ONS or labelled ‘National Statistics’. The arguments here are largely obvious: many of the most politically sensitive and controversial statistics are produced by the major policy departments of government and based on administrative data crime, education, health and social security records not statistical surveys. The new arrangements must recognise this or risk public confidence in such statistics being reduced rather than enhanced”.