Select Committee on Public Accounts Minutes of Evidence


Examination of Witnesses (Questions 20-39)

DEPARTMENT FOR TRANSPORT

9 JUNE 2008

  Q20  Mr Mitchell: So you wanted the savings to achieve Gershon's targets and you therefore overestimated.

  Mr Devereux: We certainly anticipated making savings from this towards our Gershon target.

  Q21  Mr Mitchell: Was it plucked out of the air?

  Mr Devereux: No, we spent six months working carefully in a support services review and then had further work done by Deloittes to validate the savings. The extent of the potential savings were not plucked out of the air. Let us just stay with the overall story on Gershon because it is quite important in terms of where the department is coming from. The first paragraph of this Report explains that we were set a target of £785 million to be saved in 2007/08, half of which was to be saved in cash. That is £390 million of cash savings were required out of the Department. We published our annual report just a couple of weeks back. We have actually managed to achieve £640 million of cash savings.

  Q22  Mr Mitchell: The estimates must have been too high. I see in paragraph 1.8 that "the Department delivered a design blueprint but could not agree a common set of underpinning business processes, and the resulting level of customisation of different processes for individual agencies—some of which was necessary for the operation of shared services—increased cost and complexity".

  Mr Devereux: I am not sure I am following the argument. When we started this, on the basis of quite careful work to do with benchmarking across the private sector in particular and benchmarking within the department, we sought to assess the level of achievable savings. That is where the £20 million comes from. It is true to say later on, as we went through this, that we have difficulties in delivering it, which is what the National Audit Office is recording, but the savings themselves were not plucked out of the air.

  Q23  Mr Mitchell: You were predicting savings and then being sloppy about the planning of them. Why did you not have a competitive bid for it? Why was the work entrusted to IBM? Were you mates? Was there some kind of feeling of friendship for IBM? Admiration?

  Mr Devereux: No. Let me come back to the point the Chairman has already identified. We deliberately set off on this with an ambitious timetable. It was our assessment that three things were true, one was that we were intending to build this on the systems which DVLA had already started on and they themselves were being supported by IBM; secondly, because we wanted to get on with it and we were being advised it was possible to do this in the space of a year, not in the space of two years. The fact that we already had a contract—which had been won in competition by IBM, let us be clear—and had the flexibility to do this, suggested to us that actually the alignment between business processes which had already been defined in the DVLA, and a contract and a contractor which had perfectly properly the ability to do this work, meant that we were able to deliver the one year target we originally set ourselves.

  Q24  Mr Mitchell: You also thought you could trust them and you were wrong.

  Mr Devereux: I do not believe in trusting suppliers.

  Q25  Mr Mitchell: They did not tell you you could have got more competitive rates and that you were eligible for them because of the increased volume of work.

  Mr Devereux: I perfectly understand, and so does DVLA, the structure of this fairly complicated contract. It provides for the DVLA on 12 months' notice to get the best available rates.

  Q26  Mr Mitchell: So it is all the fault of the DVLA.

  Mr Devereux: No. I am just describing the contract. The contract describes that if the DVLA or the Department—call it the Department if you wish—gives IBM 12 months' notice we get the best rates, unsurprisingly, because they have time to plan.

  Q27  Mr Mitchell: Paragraph 1.8 tells us that the "Driver and Vehicle Licensing Agency and the Driving Standards Agency were themselves insufficiently prepared for implementation".

  Mr Devereux: Yes.

  Q28  Mr Mitchell: Yet you are trusting them to get the best rates.

  Mr Devereux: There are two separate sections of the Report. There is one about the merit or otherwise of choosing the IBM contract which, as I said, was because of the timetable that we chose to go with and the fact that the contract enabled it. I have explained why the rates were not as low as they might have been because we deliberately chose not to give 12 months' notice as we intended to finish this in 12 months, therefore we were using higher rates. The second point that you are making about the extent to which the DVLA and DSA were not prepared. If you read it, it is about their actual internal preparations: did their staff understand how the system would work and about the organisation, it was not about the IBM contract.

  Q29  Mr Mitchell: You are answering very fast; it is all a bit glib to me.

  Mr Devereux: You are asking a lot of questions and I am trying to answer each of them.

  Q30  Mr Mitchell: Paragraph 1.16 tells us that, "the Department's use of the framework agreement to commission work and manage IBM were not conducive to controlling costs. For example, the Department did not take advantage of cheaper day rates that were available under the framework agreement for bulk work due to the piecemeal way in which the work was commissioned." Did you have indication about this from IBM, that you could have got it cheaper?

  Mr Devereux: As I have sought to explain, this sentence is basically saying: had we given them 12 months' notice we would have got lower rates. It is not saying that the way we went out to contract it we somehow ended up with higher rates than we should have got for something done at that notice. There is a distinction here between the extent of notice, which I am recognising we did not give for the reasons I have already given the Chairman (about the ambition that was behind this Department trying to get on ahead of Whitehall to deliver this properly), and the reality and the structure of the contract.

  Q31  Mr Mitchell: The whole thing was based on the DVLA and it seems apparent from the Report that they themselves were not ready, they were not prepared. Paragraph 1.28 tells us that and paragraph 1.4 tells us that a "supporting IT system should be built on the existing processes and systems at the Driver and Vehicle Licensing Agency and the Driving Standards Agency, but take account of the needs of the wider Department". It seems to me the concentration of work and the central focus being on the DVLA actually got you into somewhat of a mess.

  Mr Devereux: Perhaps I can take you to paragraph 1.12. That describes in some detail exactly what was at DVLA pre-existing this decision and what was not. It basically says that the DVLA and the DSA used an older version of SAP to provide finance, some basic human resources (by which they mean an internal database of staff) and some procurement functions. That is a lot less than we have now delivered in the Shared Service Centre. To give an example, there is no employee self-service in that pre-existing DVLA world; there is no way of accessing that pre-existing DVLA world. It is correct to say that we built on them because they had already implemented a SAP system; it was already capable of doing the right thing in terms of the way in which you use invoices, but it was only part of what we were trying to get to.

  Q32  Mr Mitchell: Paragraph 1.12 tells us, "The newer generation SAP Enterprise Resource Planning system could be achieved either by upgrading the old SAP system or replacing the latter with the newer version, and the Programme Board elected to adopt the latter option. Best practice would have been to review the business case at this stage to assess the impact of this change". Why was that not done?

  Mr Devereux: The short answer to that was that the absolute difference in costs between a licence for the new system and the old system is quite small. As the Report goes on to say, some of the things which, having got the new system, were then possible which may themselves have added to cost is a separate issue and would not have come up at that point. When it says "elected to adopt" it makes it sound as if they did this on a whim. There is a very substantial piece of analysis that the Programme Board did to decide which of these two courses was the best one to go for, and that is the one they based their decision on.

  Q33  Mr Mitchell: Since that was signed off in January 2006 it must have been clear by then that some of the economies you had expected from the original were not going to occur and the costs were going to be increased.

  Mr Devereux: What is going on in January 2006 is the Board re-visiting what they decided in April 2005. The April 2005 decision is, by definition, an outline business case; it is before any detailed design. By the time you get to January 2006 we know the answer on the design, we know that we have chosen SAP Enterprise Resource Planning system. At that point we still have a project which is projected to have a £37 million present value.

  Q34  Mr Mitchell: What is the situation now? Have you become addicted to IBM? You are dependent on their consultants and their consultants will be withdrawn, and yet they are a substantial proportion of the teams developing the system. What happens? Can you work without them?

  Mr Devereux: IBM are significant suppliers.

  Q35  Mr Mitchell: IBM are a profit making company.

  Mr Devereux: Yes, I know. I do not imagine you would like me to be running a computer company, so we actually buy in people on the basis of a contract which they have tendered for. They have delivered, through the DVLA, for example the ability to do your tax on line, connecting it up with the MoT test; they have delivered the ability for the police and the courts to directly access the driver licence database; they have delivered the ability to make the transactions at the post office more secure. All of these improvements in customer service which the DVLA are offering their customers are based on the work they have been doing with IBM. This is a long term project and that is where it comes from.

  Q36  Mr Mitchell: I think you are too trusting of them.

  Mr Devereux: I am not going to accept that I am too trusting of them. The Report correctly identifies that we are approaching a point where we have a choice about whether to exercise a break in this contract. The Board of the Department had a long conversation ten days ago about exactly what was already on the table in terms of improving the offer. I have offers on the table both in terms of the rates for this contract and for the way in which they deal with phase six, on which Barbara wants to say more.

  Q37  Mr Mitchell: This system, the common services, applies to six, later increased to seven, executive agencies. I wondered from a point of view of Government re-organisation—which periodically come along when agencies are transferred from one department to another with the whole packaged shuffled round—how are they tied in now? This diminishes the flexibility of the system because taking common services they are all tied in to you whereas a future government might want to shove them to another department.

  Mr Devereux: Doing nothing because there may be government change is not an acceptable option either. The reality is that the hardest thing in this shared service has been to get each of the agencies to understand and define the business processes they operate. Having so defined them they will be so much better able to move almost anywhere else. The hardest thing in this is actually getting those definitions done in the first place which is what they have now done.

  Q38  Mr Bacon: Mr Devereux, there is nothing odd, strange or difficult to predict that there will be an issue about defining business processes accurately, is there?

  Mr Devereux: I do not think so.

  Q39  Mr Bacon: It is pretty normal in taking an organisation and trying to introduce a computer system to it; it is a standard thing, defining business processes accurately before you make any further progress. I do not deny that it can be quite difficult to do it accurately, as you have just said, but the fact that you have to do it and you have to get it right is predictable.

  Mr Devereux: It is, yes.



 
previous page contents next page

House of Commons home page Parliament home page House of Lords home page search page enquiries index

© Parliamentary copyright 2008
Prepared 16 December 2008