Inflation

Lord Burns Excerpts

Monday 1st July 2019 Lords Chamber

[Lord Burns (CB)]

My Lords, I too thank the noble Lord, Lord Forsyth, for securing this debate and chairing the committee so effectively. I found the inquiry rather a strange experience to begin. I had spent almost 20 years in the Treasury worrying about how to control inflation, yet in this inquiry we were deep in the detail of measuring inflation down to a few decimal points. In the process, as has been mentioned, we became aware of a series of quite surprising events that cast doubt on the Government’s various measures of inflation. I should like to develop some of those concerns.

I well recall that the main governance for the RPI until the mid-1990s was the existence of the RPI advisory committee, which consisted of some officials and a number of stakeholders, including the trade unions. The noble Lord, Lord Lea, mentioned that he was for many years a representative on that committee; there were also business representatives. My recollection is that the advisory committee found it difficult to accept any change that made a significant impact on the inflation rate, one way or the other. Indeed, there was a strong and continuing concern to uphold confidence in the RPI measure of inflation, for the reasons he mentioned.

It turns out that the RPI advisory committee did not meet between 1995 and 2007. I was somewhat surprised to learn this. It was charmingly referred to by the ONS in its evidence to us as a period of “no governance”. Then in 2007, we had the Statistics and Registration Service Act and the introduction of the RPI protocol. This required the ONS to produce a monthly figure for the RPI and introduced the formality we have heard about: that the Chancellor had to give his consent to any fundamental changes judged by the Bank of England to be materially detrimental to the holders of relevant gilts.

Paradoxically, this legislation, designed for the worthy purposes of increasing the independence of the statistics authority and improving the governance of the RPI, turns out to be a contributory factor in the loss of confidence in the RPI as a measure of inflation. In my mind, this stems from the asymmetric treatment of changes that are detrimental to the holders of gilts, as opposed to changes that are to their advantage and to the detriment of others. This is set out in the statute.

As we have heard, where a fundamental change is seen as materially detrimental to the holders of gilts, the Chancellor, with advice from the Bank of England, has to decide whether the changes should go ahead. By contrast, if changes are beneficial to the holders of indexed gilts, the Bank of England is not required to take any action. This asymmetry became evident in 2011 and 2012, when there was a change in measurement of clothing prices, as we have heard. I argue that the statistics authority then made some quite serious mistakes.

The effect of the change was, as we have heard, an unexpectedly large increase in the clothing component of the index and an increase in difference in the growth of the RPI and CPI to around 0.8% a year, instead of 0.5%. I stress that, in the evidence we received, there was a lot of criticism of the change, along with claims that it had not been tested before implementation. This was the first mistake.

What I conclude to be a second mistake followed, which was not to undo the change and to go back to the previous arrangements reasonably quickly, when the emerging problems became evident. It became clear that the statisticians were influenced too much by worries that it would be judged a fundamental change that was materially detrimental to gilt holders. The one-sided nature of the protocol meant there was no requirement to be concerned about the original change, which had materially advantaged gilt holders. Instead, the options were studied and the focus switched to the weighting system and horrendous technical debates about the merits of different methods of compiling the two indices. This response is a classic case of the best being the enemy of the good. Reversing the clothing changes would not have removed the whole difference between the two measures but would have dealt with it in part. Reversing it quickly might also have been seen as a correction and not a fundamental change.

There followed what I think we all agree was a third mistake: the decision to maintain the RPI in its current form, but to declassify it as a national statistic and consider it a legacy measure, with no further improvements to be made. This was astonishing, because it was evident that the RPI would be in place in contracts for many years, both for gilts and pensions. The committee raised the question of whether admitting that this statistic is flawed, but refusing to fix or maintain it, leaves the authority failing in its statutory duties.

Another related governance aspect of this story worries me, which was emphasised by the noble Lord, Lord Darling. The authority admitted that it had been reluctant to propose a change to the Bank of England when there was a significant risk that it would be told that it was a fundamental change likely to go to the Chancellor. This fails to follow what is set out in the legislation. I have some experience of public bodies, where the framework for their independence is set out in statute but there is a requirement to obtain the agreement of Ministers on a limited number of occasions. My interpretation is that it is for the public body to take a view about changes that should be made on professional grounds and not to shrink from referring them to Ministers for their approval, when required. In this case, it is not for the statistics authority to seek to guess the Bank’s response before deciding whether to propose changes; the decision should be taken on professional statistical grounds. It is then for the Bank of England to decide the materiality and potential detriment and for the Chancellor, in turn, to take a view on whether the proposed change should go ahead.

The committee has come forward with a sensible and workable set of proposals to try to get us out of this stand-off. At the same time, we should reflect on aspects of the governance of national statistics. As I said, the drafting of the legislation is unhelpful because of the asymmetric treatment of gilt holders and other stakeholders, not least those saving through government saving schemes. Even taking the legislation at face value, surely it is possible to make changes necessary in the light of changes in markets and product innovation without them being classed as fundamental and so that, when a mistake is made, repairing it is seen not as a fundamental change but as a tiny correction.