Previous Section Index Home Page

That was the point that I was trying to make to the Financial Secretary. The example that I gave was that we have the Bank of England basing interest rate decisions on a consumer prices index inflation rate of 2.7 per cent., while the reality for some people, especially pensioners, is an inflation index almost as high as 9 per cent. Is it really surprising that trust is undermined when personal experience seems to bear no relation to official figures? An ONS spokesman said recently:

That reasoning is not good enough, because it suggests that the real point of statistics is not to give the public the information that they want but simply to produce the lowest possible number on which to base policy decisions.

Public confidence stems from clarity and simplicity. Form must follow function in our national statistical architecture, not vice versa. But the form is not just window-dressing. The best impartial statistics in the world are useless if they are not generally believed. The difficulty of the system proposed in the Bill is that it is neither clear nor simple, and its products will be no more likely to be believed by the public than those that are in place at present.

My first point is that decentralisation of the kind preserved in the Bill does nothing to encourage public confidence in statistics produced by the Government Statistical Service, particularly those not designated as national statistics. As the Committee’s inquiry found recently, the public are unlikely to distinguish between different sources of data and the varying standards that apply to each. The scattered delivery of statistics by heads of profession who are based within Government Departments may indeed deliver advantages by placing the statistician closer to users, suppliers and policy makers while sharing expertise across Government, but the question remains over who the user of statistics really is: the Government, the public or both?

8 Jan 2007 : Column 73

The statisticians of the Government Statistical Service are consistently double-hatted, because they prepare statistics for public consumption and provide the data which inform and monitor the policy of their parent Departments. The Royal Statistical Society is fairly damning about that confusion of roles—it represents a prima facie conflict of interest. As ever, no one can serve two masters. Even if the decentralised structure of UK statistics is to be preserved—the Committee heard strong arguments that it should be—the Bill needs clearly to address that conflict. If it does not, it will in every sense be a missed opportunity.

Central release of statistics by a single body is one option with compelling advantages for encouraging public trust. It would remove the temptation to use the same departmental officials both to prepare data sets for public consumption and to become involved in their subsequent use by the Departments from which they originated. The Government suggested in their response to the Committee’s recommendation in this area that

Not only is that responsibility absent from the Bill, but the reasoning behind it is circular—or perhaps just optimistic—because it merely returns us to the problems implicit in statisticians serving two masters. What happens in the event of a conflict between a civil servant’s professional duty to the national statistician and political duty to his or her Department? I find it hard to believe that statisticians themselves will be content with such a potentially duplicitous situation. Does the Minister really have confidence that statisticians will withstand pressure from their immediate superiors in favour of a duty owed to a physically and professionally remote national statistician? Perhaps the code of practice to be developed under clause 10 will have something to say on that point, but we should not have to wait and see.

Of course, many professionals, including statisticians, are quite used to the idea of reconciling conflicts of interest, and Chinese walls are common in these circumstances. Statisticians have highly organised minds, and they are presumably quite competent in keeping their various responsibilities separate. In recent years, however, the trend has been towards formalising the procedures needed to avoid conflicts of interest, through legislation if necessary. That idea was particularly apparent in the United States in the context of the Sarbanes-Oxley Act.

That is exactly what this Bill is meant to do. We are adding legislative weight to the gradual accretion of safeguards that are intended to guarantee the independence of statistics. The Bill ought to be explicit about how statisticians at the Government Statistical Service are expected to resolve a conflict of interest between professional standards and political expectations. If the Bill does not address such a central conflict of interest, that cannot help but damage public confidence. The public do not want another opaque code of practice or informal Chinese wall. They want, and the Bill must deliver, a simple and transparent division of responsibility and accountability.

My second point is that the Bill must not become a triumph of form over substance by neglecting to provide an inclusive definition of what constitutes
8 Jan 2007 : Column 74
national statistics. Not enough has been done to address the false dichotomy between designated national statistics and other official statistics. The Government’s consultation response was adamant, if slightly unctuous, about the public’s ownership of statistics, saying twice that statistics “are a public good”. If that is the case, the public’s ownership is still heavily mortgaged to the good will of Ministers.

The Phillis review affirmed the principle that

However, the Government’s response to the Committee’s concerns affirmed that

and that idea is central to the Bill. By keeping any statistics that are not “national statistics” outside the purview of the proposed code of practice, the Bill invites Ministers to ensure that their statistical output remains under their control and is subject to less stringent auditing.

It is no accident that the statistics in which the public are most interested—those on health, education and crime—are not national statistics. As we heard earlier, maintaining the existing anomalies of classification does little to encourage public faith in the system. There must be a resolution to issues such as the presentation of quarterly hospital waiting lists, but not monthly waiting lists, as national statistics.

I am also concerned about the absence of the customary “carrot and stick” inducements that might encourage Ministers to ensure that their outputs are designated as national statistics. The vain hope that Ministers will want their statistics to be awarded a quality kitemark is wishful thinking; I suspect that pigs will fly before Ministers scramble to fly their kitemarks. We need to examine methods of inducement or compulsion.

The Committee was very clear that the Government were in danger of formalising a two-tier structure with their proposals and were missing an opportunity to consolidate and simplify the existing system. We must be more rigorous about divorcing the admitted benefits of decentralisation from damage to public confidence done by ministerial control over outputs. I hope to return to that issue in Committee.

My third point is that in addition to a marked lack of certainty about delivery and content, the Bill introduces some very worrying confusion about the role of the national statistician. I am grateful that the Government responded to the Committee’s concerns about the national statistician's title, but in many ways that was the Committee’s least significant concern because it dealt with form, not function.

It seems strange that the formal separation of the executive and oversight functions is still lacking in the Bill, given that elsewhere the Government are pursuing the doctrine of the separation of powers with all the fervour of a recent convert to the gospel according to Montesquieu. Much like the staff of the Government Statistical Service, the national statistician is expected to wear more than one hat—as chief executive of the board, head of the Government Statistical Service and chief adviser to the Government on statistical matters.

8 Jan 2007 : Column 75

However, the board’s responsibility for both its own statistical output and oversight of the national statistician’s work is more problematic, which was a point made earlier. The conflation of executive functions and a role in scrutiny is a recipe for uncertainty and it is not clear how the new arrangements will improve existing oversight. At the same time, the abolition of the Statistics Commission will do little to bolster public faith in the system. The Bill has another spasm of optimism about the national statistician’s advisory role by imputing to Ministers a willingness to listen to advice rather than recognising the occasional need to compel them or—in extremis—to remove decisions on statistics from them altogether. In short, the national statistician must not be toothless.

My final point is that the ambiguous place of scrutiny in the Bill raises concerns about the residual involvement of Ministers in the operation of the board. The announcement of the board as a non-ministerial Department has made a positive splash, but if Ministers can continue to play the part of eminence grise with impunity, it remains a sop to independence. That will be of particular concern if the board remains under the auspices of the Treasury, which is the largest single producer-cum-consumer of statistics; many whiffs of Government interference in statistics occur in connection with it.

A suspicious person needs to look no further than the movement of the “golden rule” goalposts, about which we heard earlier from my hon. Friend the Member for Sevenoaks, or the scrap between the ONS and the NAO about whether to classify Network Rail’s debt as off-balance sheet. As a suspicious person myself, I have recently drawn attention to the slow drip of more than £11 billion of private finance initiative debt on to the Government’s balance sheet in the past 18 months as a result of ONS reclassification. Perhaps my hon. Friend was being diplomatic; I shall not be as diplomatic with the Economic Secretary—although I notice that he is not in his place. Did the Treasury exercise some restraint over the ONS in those belated decisions? If so, how much similar debt still awaits reclassification? We do not want to have to ask such questions of a new board.

The Bill also proposes that the Chancellor should relinquish control of the retail prices index, and I think that everyone agrees that that is well overdue. However, clause 19 provides a caveat to protect bondholders from fundamental changes to the RPI. That is rather like the Chancellor ceding control of interest rates to the Bank of England except in cases when, for example, a change would adversely affect mortgage lenders—a rather pointless gesture. In any event, the proposal is not likely to convince the public that the Chancellor does not retain the whip hand; again, the issue becomes one of public confidence. At the very least, we need to look again at whether a transfer of residual responsibility should not, after all, be made from the Treasury to the Cabinet Office, as was suggested by the Royal Statistical Society. Perhaps that would be “back to the future”.

To conclude, the Bill is as much a test of perceived as of real independence. As it stands, the Bill fails to clear
8 Jan 2007 : Column 76
the former hurdle before even getting started on the latter. The nuts and bolts will become a matter for debate in Committee, but I am concerned that Ministers are not sufficiently sensitive to the need for the public to have total faith in the Government’s commitment to both independence and oversight. If the Bill is not unequivocal about encouraging public trust, there is no point in getting started on the detail. I look forward to a reiteration of that commitment in the Minister’s reply.

6.59 pm

Kali Mountford (Colne Valley) (Lab): I disagree absolutely with the final remark made by the hon. Member for Braintree (Mr. Newmark). If we start the process by saying that the only important outcome is the public’s trust in statistics, and that if we do not achieve that we will have failed, we will have started on the wrong journey in the first place. Although public trust in statistics is a desirable outcome, it would be wrong to see it as the only outcome. That would be too heavy a burden for us to place on the shoulders of statisticians.

Some of my best friends are statisticians, although I have never been one. We could, perhaps, compare their popularity with that of politicians and journalists. I hope that at the end of this process they become slightly more popular. Politicians may not, of course, and I will say nothing at all about journalists. In any event, it would be unreasonable of us to expect statisticians to take on such a burden for the sake of our popularity, and it would not be fair to judge the Bill on that basis either.

Sir Robert Smith (West Aberdeenshire and Kincardine) (LD) What the Bill could achieve, if dealt with properly, is the separation of politics and statistics. That would give the public more confidence in judging where things may have gone wrong.

Kali Mountford: I would agree with that if I believed that what people considered to be wrong was political interference, and political interference alone. I do not think that people will feel much more confident in statistics afterwards than they did before. People perceive that there is political interference, and perception is reality. Independence from Government is often worth achieving for its own sake, because it is a public good in any event, notwithstanding public view. If we believe in decentralisation, it is a public good. That part of the debate can exist in its own right. It is not necessary to argue about whether politicians are good or bad, or about whether we should interfere in statistics.

My experience is not very great; but whatever my distance from Government, I have watched Ministers do their job, and I have never witnessed any interference or seen any Ministers wanting to interfere. Perhaps that is due simply to the fact that I work with people of high integrity. Perhaps I am lucky in that respect. I have also worked, as a civil servant, with civil servants of high integrity. I believe with my whole heart that they are worthy of high esteem. I trust them completely, and do not believe that they commit, or want to commit, improper deeds. I am certain that they would not want to do the wrong thing.

8 Jan 2007 : Column 77

I do not know of many people who start their working day by asking themselves “How can I do the public down today?” Most people do not go about their business in that way. Some do, of course. If there were not such people in the world, there would be no criminality. Most people, however, begin their day just wanting to do a good day’s work, and we should trust them. That is most people doing most work, and it includes us.

There is, however, another purpose to what we are doing. When we produce figures, they must be right, but not every provision in the Bill travels in that direction. There are other things that we could do to ensure that figures are right, and they need not be in the purview of Parliament. They could be dealt with in codes of practice that are not only the property of the civil service or the Office for National Statistics—or the Statistics Board, as it will become. There are other professional bodies with an interest in statisticians and in whether their professional integrity is working well.

We should be asking what statistics can and cannot do. We act as if we should shoot the messenger all the time and as if the statistics themselves were to blame, which they are not. The figures are not the problem. What we should be thinking about is how maturely we deal with the information contained in them.

We have heard some interesting comments today about what we should do when presented with figures, and how we should analyse them. We are always using the national crime statistics, for example. One aspect of those statistics—which is not really very complicated, but tends to confuse us—is the question of whether overall crime is increasing or decreasing. Generally, people just want to know whether they are a little bit safer or a little bit worse off. Can they feel safer or less safe, living where they do? It is not usually necessary for people to look at the national crime statistics to know whether they should feel safer in the town or village where they live. They can look around them and decide for themselves.

People see reports on television, and politicians arguing among themselves. Someone will say, “But knife crime is up”, and it will be true. We must be strong and secure enough to say that it is true, and that must inform public debate on public policy and what we should do about knife crime. Accepting that knife crime is up, however, does not mean accepting that more young people are carrying knives and using them. It does not mean that crime in general is on the increase. We must be able to accept that that one crime is going in the wrong direction, while other things are getting better. We must be mature enough to say, “But burglary is down, and car theft is down. Those things are going in the right direction.” Those are the sets of figures, and we must be mature enough to accept them in the round. We should not say, “You are massaging the figures, because you are not accepting this aspect of violent crime.” We should not be having an artificial debate about which is the true figure, when the truth is that one aspect of crime is going in the wrong direction and another is going in the right direction.

Mr. Newmark: The issue of public confidence comes down to how politicians manipulate the statistics that they are given. The hon. Lady is right in saying that if knife crime is on the increase, Ministers should be
8 Jan 2007 : Column 78
honest enough to say, “Knife crime is on the increase, but break-ins are in decline.” They should give all the figures in context. The problem lies in the fact that politicians tend to give the most favourable figure rather than showing the true picture.

Kali Mountford: If that were true I would think it was the problem, but the figures are published in their entirety. If it were true that only the burglary figures were published, that would be a problem. If Ministers came to the House and said, “We are keeping knife crime a secret from you”, that would be a real problem—but they do not. There may be a debate about Ministers’ response—about whether a knife amnesty was the correct response, for instance. It is right for Parliament to debate such matters, but it is not true that Ministers should not respond to a change in the number of people carrying knives.

It would be bizarre if Parliament did not have a view on whether Ministers’ response to knife crime was the right one, but it would be odd indeed to suggest that Ministers should not come to the House with a view on knife crime. It would be immature of the House not to be able to handle the fact that there has been a change in the incidence of knife crime. That is a simple fact. There is a difference between the fact that there have been statistical changes in crime, going in different directions, and—if the statistics have been published, as they have been—the issue of whether Ministers have responded correctly.

Peter Bottomley rose—

Kali Mountford: The hon. Gentleman looks confused, so I shall give way to him.

Peter Bottomley: This may be the hon. Lady’s interpretation of the argument that we are actually having about Ministers’ having information too far in advance. If I quote a headline to her—

Next Section Index Home Page