Statistics and Registration Service Bill

New clause 4 and amendments Nos. 16 to 22 would extend the full rigour of the reforms to all official statistics. We agree with the Royal Statistical Society, which stated in a letter to the Chancellor last year:"““The new arrangements should cover all official statistics not just those which are currently the responsibility of ONS or labelled National Statistics.""If statistics produced by the major policy departments on topics such as crime, education, health and social security are omitted, then this will erode public confidence, rather than enhance it.””" Our amendments would abolish the confusing two-tier system between national and other official statistics. They would give the board the authority to regulate the statistical activities of all Government Departments by applying the code of practice across all Departments. They would address one of the most significant weaknesses in the Government’s proposals. It is startling that the Bill does not oblige anyone to comply with the code of practice, despite the Government’s promise to put the code on a statutory footing. New clause 4 would remedy that by imposing a legal obligation on Departments to obey the code of practice. I am pleased to see that the Liberal Democrats’ amendment No. 44 would go in the same direction. As the Minister stated in Committee, the Bill as drafted does not give the board a supervisory or regulatory role at all, but gives it a function that is ““much softer in nature””—one of assessment and audit, not supervision. As drafted, all the Bill does is require the board to produce a code and to empower it to carry out assessments of particular statistics against the terms of that code. As Professor Tim Holt has pointed out, the Bill leaves it to Ministers to decide whether the code of practice is to apply to them, because clause 12 leaves it in their hands to nominate statistics for assessment by the board as national statistics. It is up to them as to whether the board can carry out even the limited assessment function permitted it by the Bill. That specific point—the starting of the assessment process—is addressed by the Liberal Democrats’ amendment No. 34, which would enable the board to initiate an assessment. That would provide a welcome limit on ministerial power in this area and would be an improvement on the Bill, but it does not go far enough. Only the imposition of a legal obligation on Departments to obey the code would give the board the authority to ensure that good practice is observed right across Government. At present, clause 7 gives the board the objective of promoting and safeguarding good practice, but not the authority with which to do that effectively. In Committee, the Minister stated repeatedly that he expected the board to promote the code as a model of good practice, but only if the amendments were accepted would the board have the supervisory and regulatory tools to ensure that the code is actually applied and obeyed across Government. As drafted, the Bill leaves the board only with the power of exhortation—the power to name and shame Departments that transgress. In the Treasury Committee, Dr. Fellegi pointed out that that is little different from the current powers of the Statistics Commission and status of the existing code of practice. The Government have admitted that the current structures are inadequate to restore trust in official statistics. Leaving Ministers to determine which of their Department’s statistical activities are subjected to the code and to the scrutiny of the board will leave the coverage of the reforms patchy and inconsistent. The British Society for Population Studies has described the national statistics system as having been"““applied in a piecemeal way to individual datasets””." If it is left to Ministers to decide which statistics are brought into the national statistics system and subjected to the code of practice, many important indicators could be left out of the scope of the reforms. I am informed by the Library that in the seven years since the 2000 reforms, there has been a net addition of only 25 new national statistics. Distinguished statisticians such as Lord Moser have described that aspect of the Government’s proposals as a significant flaw. If this reform is to succeed, it must, as the Statistics Commission has stated and the Minister has acknowledged, address the whole Government statistical system. There is no good reason why any lesser regime should apply to Departments than to the ONS. That makes no logical sense when it is acknowledged that the more significant problems have occurred outside the work of the ONS. As the British Urban and Regional Information Systems Association has pointed out, the social statistics produced by Departments on crime, education and health are part of the very currency of political debate in this country. If the reform subjects those to a lesser degree of rigour and scrutiny than other official statistics, it will fail in the goal set by the Government. As the Statistics Commission has sagely pointed out, that will risk public confidence in such statistics being undermined rather than strengthened by the Bill. We are not talking about trifling, insignificant exclusions: a number of important and politically sensitive figures are currently outside the scope of national statistics. I do not propose to trouble the House with the long list that I read out in Committee, but I shall pick 12 or so examples. Excluded statistics that would not be covered by the code of practice under the Bill as drafted include: figures on the end-of-month prison population count; quarterly figures for cancelled operations and time spent in accident and emergency; annual figures on NHS bed availability; business survival rates for firms still registered for VAT after three years; annual progress reports on the UK fuel poverty strategy; and armed forces medical discharges. Many important statistics from the devolved areas fall outside the national statistics framework: in Scotland, alcohol-related, health and mortality statistics on various topics and cancer audits and waiting times; in Wales, figures on detentions under the Mental Health Act 1983, and data on cervical screening and exclusions from schools. Those are significant figures that deserve impartial treatment. In Committee, the Financial Secretary based his counter-arguments in that regard on a claim that some statistics were more important than others. In essence, he asserted that not all statistics were important enough to be subject to the code of conduct. That argument has several flaws. First, if it is worth collecting and relying on a statistic, surely it is worth doing it according to standards of impartiality, objectivity and accuracy, as set out in the code of practice. Secondly, it is difficult to know in advance which statistics will be significant, and which will not. For example, on Second Reading, the Financial Secretary referred to the egg bulletin as not significant enough to warrant application of the code of practice. As various salmonella scares have shown, however, statistics relating to the safety of food, including eggs, can be of critical importance. Thirdly, the decision as to which statistics are significant enough to merit the application of the code of practice is simply too important to be left to Ministers. In Committee, the Financial Secretary rightly placed great emphasis on the boundary between important statistics and those that he dismissed as insufficiently significant to merit compliance with the code. He said that that"““cuts to the heart of the proposals in the Bill and the concerns that some have expressed.””––[Official Report, Statistics and Registration Service Public Bill Committee, 23 January 2007; c. 207.]" He is right: it does. Under the Bill as drafted, the boundary between national and other official statistics marks the boundary of the scope of the reforms. The Financial Secretary wishes that boundary to be determined by Ministers, so that they can say, in effect, to the board, ““thus far and no further””. The Financial Secretary said in Committee that he hoped and expected that the system put in place by the Bill would evolve. Given the approach that he takes, we know who will decide the nature and pace of that evolution: the very Ministers whose intervention this whole reform is designed to reduce. That aspect of the Bill significantly undermines the good intentions of the reform and retains significant ministerial power over official statistics. In Committee, the Financial Secretary said that he felt there would be a"““stronger incentive…for Ministers to look actively at submitting additional departmental statistics for approval as national statistics where they are central to the policy functions or delivery of programmes for which those Ministers are responsible.””––[Official Report, Statistics and Registration Service Public Bill Committee, 8 January 2007; c. 33.]" If he really believes that, he lives in a happier and less cynical universe than the rest of us. As Dr. Fellegi pointed out to the Treasury Committee, the possibility of enhanced and exacting scrutiny can hardly be much of an incentive to opt into any system. Anyway, if Ministers were so keen to take on that kind of scrutiny, why are the Government so resistant to calls for the expansion of the scope of the code of practice to cover all statistics? The second assertion that the Minister advanced in Committee was that the Opposition’s approach was"““a certain recipe for rendering the board ineffective.””––[Official Report, Statistics and Registration Service Public Bill Committee, 23 January 2007; c. 205.]" That is predicated on a misunderstanding of how the new clause and amendments would work in practice. They would not require the board to carry out an individual assessment process in relation to each and every number produced by a Department. Instead, they would give the board the power to monitor and supervise the activities of Departments when they produce and disseminate statistics. We want the code to apply to people, not numbers. The new clause and amendments would transform the code from a box-ticking exercise to be carried out in relation to a particular set of statistics into a code of behaviour to guide the people who produce statistics—the statisticians and officials who produce and handle statistics at the ONS and in Departments. Assessments can be part of that supervisory process, but there is nothing in the amendments that would place an obligation to carry out an assessment on any one particular statistic unless the board considers it to be proportionate and sensible so to do. After all, the Financial Services Authority has a 10,000 page, 6 ft wide rule book that it applies to the entities that it regulates, but it does not oblige the authority to check every transaction to regulate compliance effectively. The Minister’s third assertion was that imposing the code on all Government statistical activities would lead to a disproportionate burden, but why can we not trust the board to produce a code that does not impose disproportionate burdens? It has an explicit duty to minimise compliance cost under clause 26. To make that duty even plainer in this context, amendment No. 17 would require the board to have specific regard to that duty in drafting the code and supervising compliance with it. The underlying thrust of the reforms is that we can trust the board to take crucial decisions relating to the production and release of statistics. In that case, why is the Minister so reluctant to trust it to take a common-sense, proportionate approach to drafting the code and to its impact on statistics of differing degrees of importance. In Committee, the Minister described the approach taken in this group of amendments as ““extraordinary and absurd””. Although he may disagree with the principle, he is wrong to dismiss as absurd and impractical a position that has the support of such a long list of experts in statistics. As well as Lord Moser, the RSS, the Treasury Committee, the Statistics Commission and others to whom I have referred, virtually all those who expressed a view during the consultation wanted the scope of the Bill widened significantly. That includes organisations such as the Audit Commission, the Market Research Society, Professor Denise Lievesley of the Health and Social Care Information Centre, the First Division Association, the British Urban and Regional Information Systems Association and the Greater London authority data management group. Even the Government’s own chief social researcher, Sue Duncan, has expressed a degree of sympathy on this point. The House should have some sympathy for the Financial Secretary. He is in a difficult position. His hands are clearly tied by his colleagues who want to limit the scrutiny that the board can exert over its departmental activities. His hands are tied by Ministers who are determined to keep the board away from certain sensitive statistics. If he is sincere in his wish to give genuine independence to statistical services, I urge him to stand up to his colleagues and take his reform to its logical conclusion; to allow the board to throw light on the statistical activities that Ministers would rather were kept in the shadows; and give the board real teeth to supervise the production and release of all official statistics, not just those that Ministers permit them to assess.