---

The Parliamentary Under-Secretary of State, Department for Culture, Media and Sport (Lord McIntosh of Haringey): My Lords, I know that I should not say this, but I was disappointed in this debate. The subject of the debate as on the Order Paper was so interesting and so important—the adequacy of official statistics for the determination of policy. I thought that there might be some constructive contributions as to the way in which statistics might be made more adequate for the purpose of the determination of policy. We discussed some interesting issues: the issue of public trust in statistics; a number of particular difficulties that have occurred in statistics over the past few years; the organisation of statistics; and the independence of the statistics office. All of those are interesting and to some extent important, and it is my duty and pleasure to reply to that debate, but I would have liked the debate to be more about what we could do.

I agree profoundly with the opening remarks made by the noble Lord, Lord Marlesford. Unless we have a sound statistical basis for policy—not just statistics in the sense that we have been talking about it but in my sense survey statistics as well—we are simply not going to be making the best possible decisions. I would like to have seen us debating the way in which some decisions taken in government—all governments, not just this Government—have been hindered by bad
 
23 Feb 2005 : Column 1290
 
statistical decisions. There has been no suggestion of that at all. I am sad, and I am disappointed; but I will soldier on as I always do.

On the issue of the organisation of statistics, the facts are well known, and the noble Lord, Lord Northbrook, fairly referred to them. We knew when we came into office in 1997 that the trust in statistics was low. That is why we overhauled the system and established the framework for national statistics, which became operational in 2000; and that is why we had the objective of ensuring the quality of official statistics, guaranteeing freedom from political interference in their compilation and presentation. That is why we supported the independence of the National Statistician. By the way, he may be appointed by the Chancellor, but he is appointed under Nolan principles, which were not current under previous administrations. That is why we set up an independent Statistics Commission, as well as setting up the independence Office for National Statistics. The framework set out the role of the National Statistician as both the head of national statistics and the director of the ONS. It set out the role of Ministers in relation to the National Statistician, the ONS, and the Statistics Commission.

In his characteristically distinguished contribution, the noble Lord, Lord Moser, said, "Of course you have to take into account the statistics that are produced by departments as well"—and that is true. He made some helpful suggestions about those individual department statistics and the need to strengthen the Statistics Commission. The noble Lord, Lord Oakeshott, made some helpful comments on that as well. The most important point that he made, which is a criticism of the Conservative proposals, is the need to keep analysis with data collection. I am somewhat of a cookbook statistician myself. All of us involved with survey research are pseudo-statisticians rather than the real thing—except when we read the classic text by the noble Lord, Lord Moser, I think from 1953. It was my leading text.

We have proposals from the Conservative Party that in many ways echo the changes that we actually introduced five years ago. They say that there should be a national statistics office established by statute; that is what the Statistics Commission says and it has to be taken seriously. The Conservatives want that office to report directly to Parliament, about which there are difficulties. They want it to separate the collection of raw data from the analysis of all available official statistics, and to publish analyses and reports based on them, with a separate social and business surveys office headed by the Registrar General.

There is nothing very radical about that. Everything said in this debate will certainly go into the review envisaged by the framework for national statistics in 2000. It was said that there would be a revision in five years—in other words, during this year. That revision will take place. The Royal Statistical Society has already been involved. The Statistics Commission has already made its suggestions, which have been quoted today. I want to take the whole of the debate as evidence for that
 
23 Feb 2005 : Column 1291
 
review, and certainly the contribution of the noble Lord, Lord Moser. To that extent, I very much welcome the debate.

The second issue discussed was public trust. I was puzzled by the noble Lord, Lord Marlesford, saying that there was survey evidence that people did not trust official statistics. There is; there is survey evidence about distrust of all sorts of things—politicians, journalists, estate agents and so on. It is a worrying fact, because we wish statistics to be trusted. However, in this Chamber, surely our concern should be above all with whether it is objectively true that the statistics are to be distrusted, not whether people think so. The noble Lord simply said, "People don't trust them and I think that they're right", but he did not give any evidence for that assertion. People ought to trust statistics; I think that they are trustworthy. I am prepared to debate that issue at far greater length than I have available.

The third element in the debate was criticism of particular statistical issues. I am happy to respond to those, but I shall not take them in the order in which they arose. I shall deal first with pension contributions. David Willetts drew blood; there is no doubt about it. There had to be very significant revisions of the estimates of pension contributions. It is common ground that pension contributions are difficult to measure. Successive legislation made the pension industry more complicated—products changed, but old products remain in the market. There is no consistent accounting in companies themselves. After all, the information comes from companies; it is not collected from the public sector. There is no consistent way in which group personal pensions are regarded as company products by some insurers and as personal pensions by others.

The problem is enormously difficult, and the Office for National Statistics had to withdraw the figures for a considerable period, which is an embarrassment. The questions then have to be asked, "Does that matter? Has it actually made any difference to public policy?". Throughout the passage of the then Pensions Bill through Parliament, I did not hear any suggestion that the volume of pension contributions had affected policy in the area.

Baroness Noakes: My Lords, I have a question before the Minister leaves pensions. I also raised with him how government Ministers used statistics in a way that was not consistent with their underlying worth at the time, as noted by the ONS. The issue was how politics and statistics were pulled together. Will the Minister say something on that?

Lord McIntosh of Haringey: My Lords, that is not the subject matter of the debate. I shall come to the point with particular reference to health, but I shall press on to road maintenance, which was referred to by the noble Lords, Lord Northbrook and Lord Oakeshott.

Let me assure the noble Lord, Lord Oakeshott, straightaway that the initiative was that of the ONS. It identified a double count in the current expenditure. That is due to the difference between resource accounting and the national accounting frameworks. Before the
 
23 Feb 2005 : Column 1292
 
change the national accounts had included both a depreciation charge, calculated by the perpetual inventory model, and the capital maintenance charge, as reported in the Highways Agency's accounts. You could not have both, so one had to be left as revenue, and the other had to be converted to capital. That will all be explained in a document to be published on 28 February. I am sorry that it is not possible to alter the publication of ONS documents to suit the convenience of this House. Even if we were able to influence the ONS—we are not, because it is independent—we would not seek to do so.

In any case, the double counting had to be remedied. It is a matter for the Office for National Statistics. The press release in which it described it has been published, and a website is available, but I certainly cannot read it out. I assure the noble Lord, Lord Northbrook, that everything is entirely transparent. The role of the Treasury in the matter was only as the provider of data for the reanalysis, not in either instigating or influencing how the inquiry was carried out. I noticed that he quoted last Saturday's Financial Times; he will find a much more perceptive article in the Financial Times today that explains the matter.

I move on to health statistics and public sector productivity. I have said on many occasions from this Dispatch Box roughly what the Health Secretary said and was criticised for saying by the noble Lord, Lord Brooke. The method is rubbish; not only that, it is counter to what common sense would tell us. The example that I have always given is that, if you reduce class sizes and have better quality information, it shows up as no change in a school's output and therefore as reduced productivity. All in-patient hospital procedures had equal weight so, whether the problem was an ingrowing toenail or a heart and lung transplant, they counted the same. If you take steps to keep elderly people in their own homes with a better standard of life, that showed up as reduced productivity compared with putting them in expensive residential care.

The method was always rubbish, and is rubbish now. That is why we have appointed Sir Tony Atkinson to look at it. His report came out on 31 January. It is on that that we should rest our case. The noble Lord, Lord Northbrook, is entirely right to quote Sir Tony's conclusions, but I am sorry to say that the conclusions drawn by the noble Baroness, Lady Noakes, are way wide of the mark. We have been doing exactly the right thing with something that is simply no good.

Oddly enough, the paucity of examples of bad statistics has been a feature of the debate, so I shall conclude by saying something about the necessity for an independent Office for National Statistics, and the fact that it has to disagree from time to time with the Government and the National Audit Office. Such a disagreement is not a criticism either of the Office for National Statistics or of the National Audit Office. Above all, the role of the Office for National Statistics is to reconcile what data we have with international accounts. Those international accounts are, in turn, influenced by international accounting conventions. They follow well documented standards used by the IMF, the OECD and the European Union, and they
 
23 Feb 2005 : Column 1293
 
are compiled on a legal basis following the 1996 regulation from the Council of the European Union. That is why there are differences between statistics produced by departments for different purposes. On occasion, there are differences between the statistics produced by the Treasury for its own purposes, but there is nothing sinister about that. It simply means that the national statistics, which are the function of the Office for National Statistics, are a peculiarly difficult but very worthwhile area.

I have to say that if this was a Conservative debate in the run-up to a general election, it has not drawn blood.

---