Statistics and Registration Service Bill

My Lords, this Bill is to be welcomed because it could, if suitably amended, enable people to have a basis on which to decide when and how far to place their trust in statistical information. Statistical information, like other information, can be based on better or worse evidence that is communicated more or less adequately; hence more or less open to check and challenge and to the intelligent placing and refusal of trust. Where we lack expertise—most of us lack expertise on most, if not all, series of statistics—we need indirect, but reliable evidence that can be used to distinguish the more trustworthy from the less trustworthy. We also need that evidence to be carefully communicated. Only so, can we move from a world in which we suffer policy-based evidence to one in which we might attain evidenced-based policy. Therefore, I welcome the fact that the Bill not only assigns responsibility for the quality of statistics, but also for their communication to a board that is more independent of government departments. Clause 7 assigns to the board duties both to promote and to safeguard the production and publication of official statistics that serve the public good. Clause 7(2)(a) and (b) deal with duties to inform the public about economic and social matters and duties to assist in the development and evaluation of public policy. Knowledgeable commentators have wondered whether the board will have enough independence to secure public trust, as it will combine executive responsibility for the production of national statistics—although not for official statistics produced by other public bodies—with powers to assess the standards to which all statistics are produced. Although the point is a strong one, I believe a case could be made for the adequacy of this structure, provided the board had not only the power to assess and to communicate the compliance of official statistics with the code of practice that it is to publish, but also the requirement to do so. However, as I read the Bill, a requirement to assess official statistics for designation as national statistics has to be triggered by a request from an appropriate authority; that is, from a government department, including departments of the devolved administrations. Does it then follow that the board will not assess official statistics where no request is made by the relevant Executive? I would be grateful if the Minister could confirm whether that is the intention behind Clauses 12 to 14. The board being required to communicate an assessment of the adequacy of all official statistics to the public would enable public judgment of those statistics; the duties of the board to communicate could not then be met by self-certification cooked up according to some in-house recipe. However, if no assessment is required, the public will be left without systematic evidence of the adequacy of official but non-national statistics. They will still lack a proper evidential basis for placing and refusing trust intelligently. As we all know, this is not an abstract consideration. Plenty of statistical series are regularly published which are not fit for purpose. I mention just one: the use of the number of passes at grades A to C at GCSE. We sometimes hear complaints that people compare apples and pears; we have a system that compares peanuts with melons. Trust requires not only good statistics, but good communication. This is not just a point about pre-release, although that is a serious set of issues. There are no real short-cuts for communicating adequately with the public. The many fashionable but one-sided activities that go under the names of disclosure, dissemination and transparency are simply not enough: they cannot ensure that others come to know anything at all. They may provide antidotes to secrecy, but are no guarantee of intelligible communication to various audiences. The board’s success will hinge as much on the quality of its communication with the public as on its reputation for independence and professional competence. Both competence and independence could be exemplary—although, as we know, there are difficulties with independence and merging the two—but if communication fails the public will rightly withhold their trust. I say ““rightly”” not because I fantasise that trust requires proof—where proof is available, trust is redundant—but because trust is no mere attitude, as opinion pollsters suppose, but a matter of evidence-based judgment, and stymied if the evidence is not even communicated. The best way to gain trust is to communicate what the case is, how something is done and to stress the limitations as much as the merits of the evidence offered. One further aspect of the communication that the Bill requires may be a source of some concern; the noble Earl, Lord Northesk, has already commented on this. In the Bill, 16 clauses are devoted entirely to issues of data protection and information sharing; that is a hefty proportion. Statistical information about populations and social and economic trends is, of course, based on anonymised information about individuals, so concerns about privacy are unavoidable. But I will not be the only Member of your Lordships’ House who has concluded, perhaps reluctantly, that our data protection legislation generates more problems than it resolves. That is why we are faced with16 clauses. Briefly, the trouble arises because our data protection legislation seeks to secure privacy by controlling or sequestering types of information defined as personal or sensitive, rather by regulating the ways in which we communicate information of all sorts, including that which is neither personal nor sensitive, as traditional conceptions of confidentiality sought to do. For example, Clause 36 contains some of the consequences of relying on this conception of personal information: information identifies a particular person, and so counts as personal and may not be disclosed by the board, "““if the identity of that person— (a) is specified in the information, (b) can be deduced from the information, or (c) can be deduced from the information taken together with any other published information””." This criterion is both wide and vague. The amount of information in circulation about most persons provides a huge corpus for inference, which often and unavoidably enables the identification of persons and aspects of their lives rightly seen as private matters. There is a great deal to be said on this matter, and I expect that your Lordships will be pleased that I shall not say it. However, I hope that the Bill will not add to the great morass of data protection confusion.