Government: Independent Statistics

Part of the debate – in the House of Lords at 4:54 pm on 15th June 2006.

Alert me about debates like this

Photo of Lord Northbrook Lord Northbrook Conservative 4:54 pm, 15th June 2006

My Lords, like other speakers, I am grateful to the noble Lord, Lord Moser, for initiating this debate. I claim no special interest in this subject, except as a director of a fund management company. I see the way that economic statistics move the markets. So, from that viewpoint, their accuracy, as far as that is possible to achieve, is vital.

There has been cynicism about statistics for many years. Andrew Lang, the prolific Scottish novelist and expert on mythology and folklore, is noted for saying:

"An unsophisticated forecaster uses statistics as a drunken man uses lamp-posts—for support, rather than illumination".

Yet statistics, not least of all, official statistics, matter. They are a necessary input into vital policy decisions by Governments, companies and individuals. To take just one example, the Bank of England's Monetary Policy Committee, when making its interest rate decisions, relies heavily on a battery of survey evidence and official data—even though those data are inevitably lagging indicators.

Unfortunately, public confidence and trust in official data have deteriorated in recent years, at the same time that users of statistics have become more demanding. The much-respected head of the Nuffield Trust, Sir Denis Pereira Gray, said at a recent seminar at the Cass Business School that the crisis of confidence in national statistics is getting worse. As the Financial Secretary to the Treasury, John Healey, stated at the same seminar, there is a need to build and reinforce public trust in the statistical system, which will take a long time.

Suspicions remain, as other speakers have mentioned, that there is ministerial interference in the classification and presentation of official statistics, and the Treasury, as I will demonstrate, is a prime culprit in that exercise. The first example was the altering of the goalposts in determining the golden rule. Its refusal to classify Network Rail's liabilities of some £17 billion as public sector debt was incorrect. Also, the extent of private finance initiative liabilities is understated, according to Capital Economics, by some £25 billion, thus ignoring the requirements of UK accounting standard FRS5, which seeks to ensure transparency in corporate accounting for off-balance-sheet items. Also the public sector unfunded pension liability has, according to the actuaries Watson Wyatt, been understated by some £230 billion and the local government pension shortfall by some £71 billion, because too high a discount rate has been used.

So it is ironic, but nevertheless welcome, that the Chancellor, as Minister for national statistics, has made moves to improve the independence of statistics. Last November, he announced that the Government intended to,

"legislate for . . . the creation of an independent Governing Board for the Office for National Statistics, with delegated responsibility for meeting an overall objective for the statistical system's integrity".—[Hansard, Commons, 28/11/05; col. 78.]

In March this year, the Treasury launched its consultation document. I found the document a bit of a curate's egg; it has some merit. One promising proposal is to make the ONS a non-ministerial department directly accountable to its independent governing board and Parliament. Professor David Rhind, chairman of the Statistics Commission, stated at the recent Cass Business School seminar that the key points about the board were its two roles: the oversight of the Office for National Statistics and scrutiny of all government statistics. The board also has an important role in reporting to Parliament.

However, I am concerned that unless the powers given to the governing board are considerable, it may not be able to resist ministerial interference in the ONS's statistical output. At a recent seminar, Professor Adrian Smith, principal at Queen Mary, University of London, had concerns about how the board would function. Would it have executive powers, a monitoring role, or both? I put the same question to the Minister.

At the same seminar, John Pullinger, chairman of the Royal Statistical Society national statistics working party, thought that the board could play a leading role in getting departments to work together to create a single statistical system but that would be successful only if it was non-executive. He believed that executive responsibility, as the noble Lord, Lord Moser, stated, should remain with the National Statistician, who should have a statutory duty to lead the ONS and to co-ordinate colleagues across government departments so that there is a single agreed picture of what is going on in society and the economy.

One move that may improve confidence in the ONS's output is for the Treasury to relinquish its role as the ONS's parent department. Even if the ONS becomes a non-ministerial department, as proposed in the document, it is likely that there would still be Treasury departmental involvement. An ideal could be an organisational structure similar to the National Audit Office or the Electoral Commission, answerable to Parliament rather than to the Treasury. Or, as the noble Lord, Lord Moser, stated today and in the debate of 23 February last year, the ONS could report to the Prime Minister, through the Cabinet Office, where its predecessor was until 1989. However the document is silent on this subject.

In my view, it would be foolish to see independence as a magic solution: scrutiny is also key, either through a new body or a strengthened Statistics Commission. A scrutiny body should be able to influence the quality of official statistics over time. It would help to identify a gold standard applying to key statistics and it would drive up the quality of other statistics by establishing high expectations.

The document's biggest failure, in my view, is the relative discussion of the data issued by departments. These include such politically sensitive issues as the Home Office's crime figures and the Department of Health's hospital waiting times, or subjects such as public sector productivity, where early experimental data from the ONS showed a truly dire picture.

These departmental data are especially vulnerable to ministerial spin on release. The big question is: what can be done about these figures? I ask the Minister whether the Government would consider the radical proposal to centralise all data compilation and delivery in the ONS, out of government departments. Alternatively, a half-way house would be to transfer release procedures from the departments to the ONS, distancing them from any accompanying ministerial statements. Without that transfer of power, the notion of independence of statistics is a pipedream.