Statistics and Registration Service Bill

My Lords, it is a pleasure to speak in a debate after the noble Lord, Lord Moser, the doyen of the statistical world. Luckily, my study of statistics began in easier times than did his. I hope that the Minister will have reflected on the detailed scrutiny made in another place and that, as we proceed, will move into the centre ground of the powerful arguments there made. I shall follow the noble Lord, Lord Desai, by considering the inherent difficulty of being a government statistician. Statistics are an academic occupation; politics are not. As with all academics—economists, for example—statisticians will debate and disagree. Indeed, as professionals, they enjoy nothing so much as a conference at which, expressing the greatest respect for each other, they will demolish each and every statistical endeavour. In private, as it were, but also when in public, statisticians are aware that somewhere close by is a fellow professional with a different point of view. Statisticians know also that the figures which they produce will be used and misused to support political argument and lobbying. They may wish that it were not so, but they know that there will be angled selection when the messages from published statistics are available to be put across. It is not always manipulation; it is just choosing the most supportive figures. Therefore, Parliament needs to be cautious when it claims that, by legislation, it can create trust. The best that it can claim is that the strengthening of institutional arrangements will enable the public to believe in the best endeavours and personal integrity of those who produce the figures. Nobody has suggested that the Office for National Statistics has been acting in any way other than in the public interest. The perceived problems are external to the office: it is the relationship between the political system and the office which needs to change. The Bill was rightly welcomed in the Commons as a step in the right direction, and rightly criticised for not going nearly far enough. There was a mixed response to the proposal to change an organisation which last year reported on 29 ministerial targets. The office is committed by Ministers to, "““create a Government Statistical Service presence in each region by March 2007””—" I wonder whether that has happened—yet the ONS does not have responsibility for the decentralised Government Statistical Service, nor is it proposed in the Bill that it will undertake that responsibility. At the same time, there are targets for relocation out of London and for a one-in-seven reduction in staff. The National Statistician is a very busy person, and the present situation is a far cry from, "““a wholly separate body at arm’s length from Government and fully independent of it””." Yet that is the Bill’s objective. Is it a truly accurate definition of the intended relationship? If it is, there are outstanding issues of governance and management structure to be addressed, following the best efforts of the opposition parties in the Commons. Some time hence, one test will be whether the list of national statistics in Clause 16 then convinces the public that the objective of serving the public good is being met. Many commentators have said that a necessary condition for reaching this objective will be that the position of National Statistician is strengthened beyond that proposed in the Bill, so that she becomes the nationally recognised guardian of statistical integrity and openness. Regrettably, the Bill as drafted weakens the position of the National Statistician. First, it proposes that she become a member of the board, with the majority of the members being non-executive. Being on a public board reduces a chief executive’s room to manage the institution effectively. Having been both chairman of a public body and chief executive of a public corporation, I have no doubt that it is more efficient and effective to have a wholly non-executive board. It also contributes to the separation of scrutiny from operations. As my noble friend Lord Goschen asked, why has the mixed membership option been chosen? The Minister’s explanation is that it conforms to private and public corporate practice. Specifically whose corporate practice do the Government have in mind? Does it conform to the ONS criteria of full independence, being at arm’s length and reporting directly to Parliament? No private body is in that position. Which public bodies are? Secondly, Clause 5(3) states that proposed head of assessment will report directly to the board. It is almost as if the board is to have two chief executives. This weakening—a partial destruction of the National Statistician’s accountability—is a strange choice and is very unlikely to work in practice. Immense goodwill is needed to make it work at all. Is it a substitute for internal systems audit? If so, would not external reviews by internationally practising members of the profession serve the board better? Peer group review is common practice among academically driven professions. The third issue arises in the extraordinary language of subsections (3) and (4) of Clause 29. Subsection (3) states: "““The National Statistician may not exercise the functions of—""(a) determining under section 10 whether to adopt a code as the Code of Practice for National Statistics or to revise the Code, or""(b) determining under section 12(1) or 13(1) whether any official statistics comply with the Code””." Subsection (4) states: "““The Board may direct the National Statistician—""(a) not to exercise a particular function, or""(b) as to how he should exercise a particular function””." Now we see the National Statistician as chief executive, and now we don’t. If the board were to direct her not do something significant or direct as to how to do something equally significant, it would almost amount to a case for constructive dismissal. What do the Government have in mind? Can they really intend to deprive the National Statistician of responsibility for the code? The code is the key to the arch of public trust. It is essential that the chief executive remains accountable for professional integrity and quality control. Production is of less importance and much easier to delegate. There are other important issues. It appears from Clauses 10 and 12 that, on the day when the Act comes into effect, there will be no new-style national statistics, but only official statistics awaiting the new code of practice and compliance with it before ““official”” can become ““national””. Is this not a highly risky procedure, depending, as it will, on Ministers’ attitudes to their share of the Government Statistical Service? If the code is not firmly backed by statute, it stands a good chance of being ignored. The open position that will prevail will certainly encourage Whitehall games-playing. I await the Minister’s reply with interest and look forward to Committee, because there is much to be done if the twin objectives of independence and trust are to be achieved.