By Tony Collins
Mark Thompson was Director General of the BBC for eight years from 2004 to 2012. He was one of the highest paid in the public sector, earning more than £800,000. He’s now CEO of the New York Times Company.
When he went before the Public Accounts Committee in February 2014 he faced accusations he had mislead MPs over the BBC’s Digital Media Initiative which was cancelled in 2013. The BBC wrote off £98.4m on the project.
Thompson has emerged from the affair unscathed although he had presided over the project. Indeed he seems to have impressed the committee’s MPs who are notoriously hard to please.
In today’s PAC report on the failure of DMI, MPs appear to have preferred Thompson’s evidence over that of other witnesses. So how is it possible to come to a PAC to answer accusations of misleading Parliament and end up winning over your accusers?
Today’s PAC report on DMI criticises the BBC for:
– complacency in taking a “very high-risk” project in-house from Siemens
– spending years working on a system that did not meet users’ needs
– not knowing enough about progress which led to Parliament being misinformed that all was well when it wasn’t
– ending up with a system that costs £3m a year to run, compared to £780,000 a year for the 40 year-old “Infax” system it was designed to replace. And Infax works 10 times faster.
In February 2014 Committee chairman Margaret Hodge began her questioning of Thompson over DMI by pointing out that, three years earlier, in 2011, he had assured the PAC that all was well with the project when it wasn’t.
Thompson told Hodge in February 2011 that DMI was “out in the business” and “there are many programmes that are already being made with DMI”. In reality, the DMI had been used to make only one programme, called ‘Bang Goes the Theory’ – and problems on the project at that time were deepening but, as in many public sector IT-based projects that go wrong, such as Universal Credit, bad news from the project team was not being escalated to top management (or the BBC Trust).
How Thompson won over PAC MPs
At the PAC hearing in February 2014 Hodge asked Thompson if he had misled the Committee when he spoke positively about DMI in 2011.
Thompson’s reply was so free of reserve that it appears to have taken the wind out of Hodge.
Thompson replied: “I don’t believe that I have misled you on any other matter, and I do not believe that I knowingly misled you on this one.
“I will answer your question directly, but can I just make one broad point about DMI before then? In my time at the BBC, we had very many successful technology projects—very large projects, some of them much larger than DMI. I believe that the team, including John Linwood [then the BBC’s Chief Technology Officer], who were in the middle of DMI, had many successes—for example, digital switchover, West One, Salford and BBC iPlayer.
“I just wanted to say … everything I have heard and seen makes me feel that DMI was not a success. It failed as a project. It failed in a way that also meant the loss of a lot of public money. As the director-general who was at the helm when DMI was created and developed and who, in the end, oversaw much of the governance system that, as we will no doubt discuss, did not perform perfectly in this project, I just want to say sorry.
“I want to apologise to you and to the public for the failure of this project. That is the broad point.”
Hodge (who would normally, at a point such as this, launch her main offensive) said simply:
Usually civil servants will deny that a big IT-based project has actually failed. Many times the archetypal civil servant Sir David Nicholson, when Chief Executive of the NHS, defended the failed NPfIT at PAC hearings.
But Thompson told PAC MPs: “Here we are in the beginning of 2014—I am not going to debate with you whether or not this project [DMI] failed. I am sure we can talk about how, why, where and so forth, but it definitely failed.
“When I came to see you in February 2011, I believed that the project was in very good shape indeed. Why did I believe that? I had seen a number of programmes myself—I had been and seen parts of DMI working on ‘Bang Goes the Theory’; I knew that ‘The One Show’ had started to use elements of DMI a few weeks earlier; and I knew that a kind of prototype version of the technology had been used in the very, very successful ‘Frozen Planet’ natural history series.
“I have gone back and asked the BBC to look at all the briefing materials—I had a voluminous amount of briefing from the BBC—and there is a real consistency between the briefing I got – .”
Richard Bacon: Sorry, a real inconsistency?
Thompson: No, a real consistency between the briefing I got and the evidence that I gave. To be honest, some of this is going to go very much to the point Mr Bacon was making earlier on (about what is or is not a deployment).
Stephen Barclay: Just a second…So it was consistent, but consistently wrong, wasn’t it, because just the following month, after the consistent briefing, you were then aware that it was going to miss the key milestone? From March 2011 you knew it [DMI] was not going to hit the deadline.
Thompson: If I may say so, what I am trying to focus on at the moment is the question—I understand, given subsequent events, the perfectly reasonable question—about whether the testimony I gave in February 2011 misled you or not… My belief is that my testimony gave a faithful and accurate account of my understanding of the project at this point.
Hodge: But were you misled, then?
Thompson: Let me give you just a sense of my briefing. To be honest, there were echoes of this in John Linwood’s testimony a few minutes ago, and Mr Bacon has helped me to understand this by putting his finger on the use of one word in particular, which is ‘deployment’. This is the timeline …”
Thompson then did something civil servants rarely do, if ever, when they appear before the committee. He read from the internal briefings he had received on the project in 2010 and 2011 . Those briefings indicated all was well.
He was not even shown a draft Accenture report in December 2010 that said the elements of the DMI examined (by Accenture) were not robust enough for programme-making and that significant remedial work was required.
Thompson said that the day before he gave evidence to the PAC in February 2011 he was given an internal note which said:
“Our next release [of DMI], Enhanced Production Tools, entered into user acceptance testing this week. This release builds on the production tool we previously delivered in 2010, Fabric Workspace, and desktop editing and logging.
“We will deploy its release to pilot users in Bristol, the ‘Blue Peter’ production team, ‘The One Show’ current affairs team, ‘Bang Goes the theory’ — again — ‘Generation Earth’, weather and ‘Pavlopetri’ inside London Factual.”
Thompson had the firm impression that DMI was challenging but that the BBC was starting to deliver the system and users had been positive about the elements delivered.
Thompson said in February 2014, “Mr Bacon is right about the very bullish use of the world “deployed”, meaning, perhaps, elements that have been loaded on to a desktop but not really extensively used: that was the background to the remarks I made to you in February 2011. I am absolutely clear that at the time that was what I knew and believed about the project.”
Hodge: So you were misled?
Thompson replied, in essence, that the BBC’s business users tried to make DMI work but most of them gave up. There were tensions between the project team who were enthusiastic about DMI and the business users who, mostly, weren’t.
These were complicated, difficult issues, said Thompson. “There was a pronounced and, it would appear, growing difference of opinion between the team making DMI and the business users on how effective and how real the technology was.
“You will understand that I have been involved in a lot of projects at the BBC and in other organisations, and I can smell business obstinacy. I can smell when a business is unready, is not prepared to play ball or is constantly moving the goalposts.
“I absolutely understand John Linwood’s particular perspective, given what he was doing. He was a very passionate advocate of the project, and I understand all of that.
“In my time, which ended when I left in September 2012, I saw great efforts being made by the business—in other words, by colleagues inside BBC Vision, BBC North and elsewhere—to get DMI to work. Although there were tensions, I do not believe that those tensions, which frankly were more or less inevitable, were themselves a central and critical part of the project’s failure.”
Richard Bacon: … It sounds to me as if the people getting the business case through the main governance processes were technology and finance people. I want to be clear on what you are saying. It sounds to me as if the technology people were very gung-ho and the experience of the business people on the ground was that it was not necessarily working as well as they had been led to believe, so they probably lost faith in it. Is that a fair summary?
Thompson: “I believe that that was definitely what started to happen, certainly by the end of 2011 and through 2012. It happened for understandable reasons. This has been a troubled project…
“I thought great efforts were made in BBC Vision and in BBC North both by senior people and by some front-line programme makers to help us to get the thing to work.
“Where my perspective perhaps differs from John’s perspective – it is very easy for me to sit here and say that this project failed because some difficult programme makers refused to use it, although there may have been an element of that somewhere – is that I thought that, overall, this was a project on which there was a lot of work and effort to try to get it to work on the business side…”
Hodge asked again if Thompson had been misled when he assured the PAC in February 2011 that DMI was being used at the BBC.
Thompson: I believed it.
Hodge: You believed it?
Hodge: You believed it, but were you being misled?
Thompson: “I think that the language that the team was using, combined to some extent with the fact that I had seen what looked like a very positive demonstration of it … I had heard that “The One Show” had also started using it, and I saw a list of other programmes that were also using it. That combined with the language in the briefing led me to believe that it was being more extensively used.”
The PAC could have concluded in its report today that the BBC had misled Parliament in February 2011. But MPs used the word “misinformed” instead.
“Neither the [BBC’s] Executive Board nor the [BBC] Trust knew enough about the DMI’s progress, which led to Parliament being misinformed. While [Thompson] assures us that he gave a faithful and accurate account of his understanding of the project at that point in early 2011, he was mistaken and there was confusion within the BBC about what had actually been deployed and used.
“In its reporting on major projects, the BBC needs to use clear milestones that give the Executive and the Trust an unambiguous and accurate account of progress and any problems.”
The PAC had every right to be angry. So credible were the BBC’s assurances about DMI in February 2011 that the Committee published a report in April 2011 that reflected those assurances. It was wrong.
But there is a positive element in the failure of DMI – and that is the completely open and honest testimony of Mark Thompson.
MPs on the PAC are used to be being misled – usually by the sin of omission – when civil servants and ministers come before them. But when Thompson read from his internal briefings it was easy to see how he came to the view that DMI in February 2011 was showing signs of a success.
It was clear to MPs that Thompson had not set out to mislead.
Perhaps the moral of the story is that you can go far with honesty and openness. That’s not an easy lesson for the ministers and civil servants who have to appear before the PAC, but it has certainly served Thompson well.