Statistics and Registration Service Bill

My Lords, I applaud the Bill. I hope someone could cross-examine Len Cook on some of the issues that he is concerned about. I am sure that the pre-release question can be clarified in Committee, as could the question of how parliamentary scrutiny will work. Perhaps the Lords economic committee could give some thought to that. A more general question is how the board can track important changes in society in what people need to know. Incidentally, it is partly for those reasons that my instinct is that the board should report to the Cabinet Office and not to the Treasury, as recommended by the noble Lord, Lord Moser, in his most authoritative speech, and by the noble Lord, Lord Turnbull. There may be a role for the board in producing what might be called encyclicals, like the Vatican. For example, the question of measuring productivity in services, particularly in public services, is a very important but very old chestnut. I do not know where one now goes to for, as it were, the text of the encyclical. There are different rules about productivity measurement but, as I understand it from my days of doing economics and statistics at Cambridge and studying a book by Professor Richard Stone, national income measurement requires output, income and expenditure all to somehow line up. People ought to produce one-off papers demonstrating how what you might call statistical national income theory lines up with what we do in practice. However, all such generalisations have to be tested against practical experience of which one has some knowledge. I will take two examples from widely different fields. The first is income distribution in Britain, and the second is how we measure and use measures of national income growth rates in sub-Saharan Africa. On the first, it has been apparent for some time that there is something strange about how we read every day about the explosion of top pay, and that not showing up in genie coefficients. It is not so much the question of City bonuses not being properly recorded—at least I trust not—but there seems to be a misleading narrative shared by the Bank of England about what causes inflation at present, particularlyin the south-east of England. Obviously, property inflation comes into that, yet we have no good statistics on how many people are sharing the top£10 billion or £20 billion increases in our economy. In a debate about four years ago, I raised my concern about not being able to measure how much City earnings contribute to the growth of national income and was told that the figure was peanuts. I have been doing some counting of all these peanuts. If we take national income to be £1 trillion—£1,000 billion—£30 billion or £40 billion is obviously 3 or 4 per cent of national income. If one gets this figure wrong by £10 billion, that adds up to quite a lot of peanuts. My friend Professor Tony Atkinson, who was a colleague of mine in the 1970s on the Royal Commission on the Distribution of Income and Wealth, said that it is important that we collect proper figures for the top 1 per cent and the top 0.1 per cent to see what is happening. It is in everyone’s interests that one has confidence in that information and in information about what one might call the two ends of the income distribution. Another problem about how we report statistics relates to averages, means, medians and so on. If one’s figure for workers’ wage increases includes all the£10 billion, £20 billion and £30 billion City bonuses, one will get a different figure—and a different analysis of what causes inflation—from the figure obtained if one takes a median within the distribution. I hope that the Bank of England’s analysis of inflation will be a bit more sophisticated. It does not seem to want to know about any of this. To some extent, this goes to the problem of how the board should be constituted. I tend to be a mongrel man myself—or a mongrel dog, more likely. My experience is that I was a member for some years—wearing my TUC hat—of the Retail Prices Index Advisory Committee. Some of its members were not expert in that they were not professional statisticians, but those around the table all knew something about what was going on. I very much regret the passing of that committee. My recollection is that it was scrapped because it annoyed the Treasury by reaching the wrong conclusion about how we should measure housing costs. Someone may challenge that but that is my recollection. My second example is about GDP growth in sub-Saharan Africa. It illustrates why statistical methods across Whitehall need to be consistent. We all know that sub-Saharan Africa has gone AWOL in terms of meeting the United Nations millennium development goals. Those goals can be translated into growth of real national income per head—per capita. I say ““per capita”” because poverty is to do with per capita rates and not overall growth rates. Far too often in DfID publications, there are casual measures of 6 or 8 per cent growth but the authors never define whether the figures are of gross national income in total or even—this is of course very elementary—per head or in real terms; moreover, we are not told whether the figures are comparable with the millennium development goals or even whether, when looking at international comparisons, people are consistently measuring using purchasing power parity, which we all know is an important measure for poorer countries, or euro, dollar or pound exchange rates. In the OECD we are putting many billions of pounds a year into development but the tracking of the results of that are very much below par. That should be of concern to Whitehall as a whole, not just to DfID. If not, we will cut other budgets in Whitehall—the Foreign Office and so on—assuming that the DfID budget is being spent effectively. The measurement of its results could be improved. To underline that, it is important to see, when we look at world economic trends, what happens with population growth. That is still a bit of a taboo subject, as we found a couple of weeks ago in a debate on population. If sub-Saharan African populations grow at 3 per cent per annum, and women have a fertility rate of five, six or seven children, one must subtract the 3 per cent before one has any real growth in the per capita figures needed to attain the MDGs. There is a lack of statistical rigour in some of the analysis that we are putting out. If the new board had to write encyclicals or—to use another metaphor—set standards across Whitehall, that would cut out this rather strange dichotomy between what are called national statistics and other statistics; that was new to me today. It is a question not just of producing the statistics but of knowing whether we are happy to use other people’s statistics. Finally, we should sometimes look at the mote in our own eye when using statistics. Many of us sat through the debate last week on reform of the House of Lords. I notice that a number of people used opinion poll data—based, I trust, on different questions. Some noble Lords told us that 70 per cent of the British people want the House of Lords to be elected and others quoted polls saying that 70 per cent want it to be all-appointed. That may be another question but it is another reason why numbers more generally are not always believed; it is not just to do with numbers coming from the Office for National Statistics.