Examination of Witnesses (Questions 40-51)
MR RICHARD BROADBENT AND MR DAVID SPENCER
WEDNESDAY 26 JUNE 2002
40. I did ask a number of Parliamentary questions in an attempt to compare Inland Revenue with Customs and Excise on some of these indicators like sickness absence. I think it is interesting that in fact Customs and Excise, on the whole, performs better on these indicators than Inland Revenue: ill health, retirements and sickness absence. There is one indicator that I asked for in questions which is not in your Spring Report, and that is staff turnover. Staff turnover in the last full year given for 2001 was almost 10 per cent, which seems quite high?
(Mr Broadbent) Oddly enough, my own sense is that staff turnover is a little low for Customs and Excise now. I will explain what I mean: the greatest proportion of staff turnover we have is essentially in Central London and some of the other conurbations, and it will be a narrow measure of large numbers of people, including casuals, probably, who are coming in sometimes as a first job, sometimes because they need the work, and leaving after a relatively short period of time, and this is the sort of turnover which we have always experienced in Central London, perhaps in Central Birmingham. Outside the major urban conurbations, and excluding, for example, retirements, our levels of staff turnover are very, very low. In some areas, they are effectively zero. This makes it extremely difficult to manage an organisation when you are constantly challenged to move your resource fromyou have a problem in the South East at the moment, but actually next year it could be the North West, so a moderate degree of turnover would be quite valuable. As I say, in practice, outside the major conurbations and outside the very junior grades, the lowest clerical grades, there was a staff turnover of close to zero.
41. Can we look at some of the statistics in the Appendix relating to performance of essentially, public service agreement targets. On page 19, for example, you had the target of a 6 per cent improvement in customer service from the 1999 base line. How exactly did you get on in respect of that target?
(Mr Broadbent) It was broadly unchanged. The satisfaction index was 70.5 per cent, I think. Your target was 74.9?
(Mr Broadbent) Our target was actually to achieve a 6 per cent increase in customer satisfaction, and the outcome was broadly unchanged satisfaction, although within that, I am glad to see that the number dissatisfied actually went down, but we failed to get that target.
42. Do you know why?
(Mr Broadbent) Not completely, no, but I think there are a couple of things I would say, beyond the fact that I think the target was itself quite ambitious. If you simply ask our taxpayers, "Are you satisfied with Customs and Excise?", 89 per cent of them say yes. The actual index of satisfaction takes a mix of questions, some of them quite qualitative and creates a base line, which is absolutely fair. Direction of this target is correct, but I just wanted to make the point that underlying this are some pretty high levels of satisfaction. In the course of the last year/18 months, we did a series of things which we thought would increase satisfaction. We introduced, for example, out of hours business advice services; we improved our website; we introduced cheap rate telephone call facilities, and the amount of dissatisfaction has gone down. I think one of the things we are learning is the provision of service is not necessarily quite the same as satisfaction, and I sometimes ask myself whether, in a sense, taxpayers can be satisfied, if you like, by our providing more services. We have dealt with this target and I am sensitive to that. The target is directionally correct. I think though, that we may need to think a bit more about how we actually increase taxpayer satisfaction. I am probably personally coming to the conclusion that we will not do that by going on providing service innovations. We are going to have to get to grips with something much more fundamental, probably linked to e-technology and IT, where we change the nature of the interaction of the taxpayer with us, which is essentially what our e-business strategy is directed at, but that is a 2, 3 year timescale.
43. So are you likely to think again about this 6 per cent target and perhaps try to express it in a different way?
(Mr Broadbent) I do no think it is quite catchingI do not have a problem with the target pushing us in this direction, but I do not think we are going to meet this target. I think the real problemwe can go back, if you like, to output measurements which is easy. We can measure new services we provide, we can measure court(?) if answered. It was actually our idea, saying, "Let us go for an outcome, let us go for satisfaction", and we are learning now that it is quite difficult to measure, so I think we will try to replace it, but I would like to keep pushing to see if we can keep an outcome base rather than just going back to outputs, which are sort of Mickey Mouse, really.
44. Next, you have productivity. Target outturn of +2.5 per cent, actual outturn ¸8.95. That is a considerable shortfall. What has happened there?
(Mr Broadbent) We got the target wrong. This again was an attempt to try and create an outcome based target, and I do not want to labour the point, but I think it is sometimes underestimated how different an outcome target is from an output target, how difficult to construct and how difficult to get an organisation to meet. We used to have a series of essentially activity measures, and, in fact, the figures for those are still given in our reports. They are the unit cost of collecting tax and the unit cost of processing trade. Although the figures are not published in the Report, and perhaps we should have published them, all those measures have improved. As a matter of productivity, the unit cost of collecting tax is lower, the unit cost of processing export trade declarations is lower than it was a year ago. And indeed, I know that most of the outputs of the organisation in terms of the tax we collect and the law enforcement activities are the same or higher than last year, and you have manpower that is actually lower than last year. So they have led me to look very carefully into why this target has this rather interesting result. The answer is this: that attempt to create an outcome target is calculated by taking only a small number of our PSA measuresbecause not all of them are very easily translated into outputs you can divide by a costin fact, out of 60 measures, only 7 are included. Of those 7, 5 were actually paper based: "Did you get the returns in?" They are not, in my view, what I would call strategic outcome targets. Even then, in most of those cases, we actually met the 2.5 per cent target. There was one which was 40 per cent deficit which was debt collection, and the reason that debt collection was so far out was because of the success of our strategic drive to deal with that missing trader fraud. We have enormously increased the amount of debt we have on our books which is actually another way of saying we have uncovered more fraud and are dealing with it. It is an illustration of how complex these outcome targets are and how careful one has to be in interpreting them. We did have a debate about how far we should try to explain some of these things in this report. We concluded in the end to publish the raw data, and I think we are going to have to come to grips with this and give much fuller explanations in the future.
45. That is interesting, because on some of them you do give an explanation, but on that one, which is a stark example of where you did not meet the target, surprisingly, there is no explanation, and yet what you have given in terms of an explanation is very interesting.
(Mr Broadbent) Yes, and we do take a view to try and keep this report short. It is like mid year reports on our flowing of accounts. One always worries about sounding defensive, but I think in that case, we should have said, "A lot of evidence suggests our productivity is going up; the problem here is we have not been able to create a sensible productivity target base with outcomes", and we are going to have to think about this, because I do want to try and keep some sort of outcome based productivity measure.
46. You referred yourself earlier to the annually reported measures which appear on page 15, the unit cost measures, really, which are all printed there. There is an interesting footnote at the bottom. It says, "Not appropriate to set targets for these indicators. They are designed to provide trend information." But how can we deduce trend from this table?
(Mr Broadbent) Unfortunately you cannot, and, as I said, when I looked at the report, as one perhaps does sometimes, in hindsight, it might have been sensible to have published a series of figures. Again, we set out, as a Board, some sort of guidelines for this report. We wanted it to be clear, concise, brief, and so we say, "We are not going to get into explaining everything, we do not want trends", and I am afraid these figures dropped out.
47. But if it is going to be a trend, you have to have the figures?
(Mr Broadbent) I would accept that, and indeed, I am happy to say if we had put the trend figures in, the trend was a good one. It was the result of an overall decision to try and get a document of a certain sort. I would accept that perhaps we had not quite got it right in terms of what we published there.
48. But the comparison has to be there for the whole thing to be interpreted. Will you make sure it is there in future?
(Mr Broadbent) I would like to have a bit of a think about this document. I certainly think there are some areas where we did not provide quite enough information, and I would like, perhaps, to include some more discussion of the sort we are now having about the nature of these outcome targets. I would not like to commit myself now to saying exactly what data comes out of that, but I would accept that we have not provided enough information in that particular area.
49. Are you suggesting that you might take the unit cost information out of the report?
(Mr Broadbent) It is not a decision we have taken. I think the unit cost information may be useful. One of the difficulties with it is that it is very susceptible to changes in policy, changes in economic conditions, for example, the unit cost of tax collected is a function of the economic environment as much as our efficiency, so I am just wary that again, we might put some numbers in which then require quite a lot of explanation to interpret them, and now I am seeking to try and find a relatively small number of helpful targets. I am not saying these should not be here, I just do not feel I have reached a conclusion on that yet.
50. Just looking back, for example, on this whole question of targets that we find in Appendix 2 on the 1998 targets, you say at one point, "15 million of assets seized." That was the target, I think. "Result: 8.3 million. Many of the cases we expected to come to court did not do so during 2000-01." Why was that?
(Mr Broadbent) We have no control over the judicial system. We can bring the cases to court, but we cannot control the timetable they go through in.
51. But you said here you expected them to come to court?
(Mr Broadbent) They came to court. We misjudged the speed they would go through at. Some of these cases have taken two years because we had to bring the case to court; you have to get your conviction; you then need to have the assessment of the financial situation; then you have another hearing to assess the benefits, and that whole process has taken somewhat longer than we anticipated. I am glad to say that the pipeline has now come through, and we are going to very materially exceed that target this year, because of this flow-through effect. But we did misjudge simply the sheer amount of time the judicial process takes.
Chairman: Thank you both very much indeed.
When I appeared before the Committee on 26 June, I promised notes in response to a number of questions. These are as follows:
Question 33seizure of vehicles
The total number of vehicles seized for all Customs offences across the whole of the UK was 5,200 in the financial year 1999-2000 and 10,219 for the year 2000-01. These figures do not include HGV's as the new HGV policy was introduced in July 2001.
Total seizure figures for 2001-02 will be published in the context of the overall performance of the Tackling Tobacco Strategy later in the year.
Question 39rates of sickness absence for year ended 31 March 2002
The average number of working days lost due to sickness absence was 9.9 daysexactly the same as that achieved at 31 December 2001. This represents a 9.2 per cent reduction against the 1998 baseline of 10.9 working days.