Decision on Taking Business in Private
Good morning, and welcome to the 23rd meeting in 2018 of the Public Audit and Post-legislative Scrutiny Committee. I ask everyone in the public gallery to switch off or switch to silent their electronic devices.
Agenda item 1 is a decision on taking business in private. Do members agree to take items 4, 5 and 6 in private?
Members indicated agreement.
“Superfast broadband for Scotland: Further progress update”
Agenda item 2 is on “Superfast broadband for Scotland: Further progress update”. I welcome our witnesses from Audit Scotland, who are Fraser McKinlay, director of performance audit and best value; Graeme Greenhill, senior manager; Morag Campsie, audit manager; and Ashleigh Madjitey, senior auditor. I ask Fraser McKinlay to make a short opening statement.
Fraser McKinlay (Audit Scotland)
Good morning, members. Everyone knows that access to fast and reliable broadband is now essential for everyday life. It is transforming the way that public services are delivered and how people interact with those services. The report that I am bringing to the committee today is the third in a series of reports by the Auditor General assessing the progress that the Scottish Government and Highlands and Islands Enterprise have made in extending broadband coverage across Scotland.
You will see that the report is in two parts. The first looks at what has been achieved to date and the second covers what remains to be done and the challenges ahead. The Scottish Government met its target of providing access to fibre broadband to 95 per cent of premises in Scotland by the end of 2017. That was achieved through two contracts with BT, at a cost of £259 million by 31 March 2018, and through commercial provision. Without public investment, only around two thirds of premises in Scotland would have access to the fibre network. We found that, due to higher-than-expected take-up and lower-than-expected costs, over 60,000 more premises than planned will get access to fibre broadband.
Ninety per cent of premises that are connected through the contracts with BT are now estimated to be able to receive speeds of more than 24 megabits per second. Of course, members will be aware that the actual speeds that people experience depend on a number of factors, including the chosen broadband package and the technology used. Our analysis shows that the average experienced speeds have increased across Scotland since we last reported, but that they continue to be lowest in rural areas.
The report also highlights the difficulties that are faced in extending coverage. For example, community broadband Scotland has not delivered the anticipated benefits for rural community broadband projects. Members will be aware that the Scottish Government has committed to deliver superfast broadband, which it now means to be faster than 30 Mbps, to the whole of Scotland by the end of 2021, which is known as the R100 programme. The Government’s initial investment in the programme will be £600 million, and it represents a key part of the Government’s vision for a world-class digital infrastructure.
Our report highlights that around 376,000 premises cannot currently access superfast broadband and that, inevitably, many of those are in difficult-to-reach areas. Further investment may be required to reach all premises and it will be difficult for the Scottish Government to deliver its ambitions by the end of 2021. There is no doubt that delivering a world-class digital infrastructure is complex and involves many public and private sector bodies, so we recommend in the report that it is important for the Scottish Government to develop and publish an overall strategy for delivering its vision, including realistic timetables with targets for delivery.
As always, the team and I are happy to answer your questions.
Thank you very much. Colin Beattie will open the questioning.
Overall, the report seems to say, “So far, so good. There are some risks in the future, and we will see how that develops.” The one thing that sticks out like a sore thumb is community broadband Scotland, which seems to have had a fairly disappointing result. Paragraph 21 highlights that its main role was
“to provide financial support and advice to communities”,
which does not seem a particularly taxing remit, although the report later mentions, in paragraph 25, in-house expertise in procurement and technology. I want to drill down into what the real problem was. Was it that CBS simply did not have the financial expertise? Was it asked to do things in its remit that were outside its experience and skills?
I will kick off on that and then ask the team to come in with a bit of detail.
There is no doubt that, as you say, community broadband Scotland did not deliver as people had hoped. Everyone underestimated the complexity and scale of the challenge. In exhibit 4, on page 15, we set out some of those difficulties. We recognise that some of them were outwith community broadband Scotland’s control, but in particular there were issues with the skills that were available in CBS to understand the complexity of the issues and with the way in which it communicated with communities. We heard strongly from some communities that they did not feel that CBS communicated as well as it might have done.
Ashleigh Madjitey might want to say a wee bit more about the issues.
Ashleigh Madjitey (Audit Scotland)
We spoke to three communities about the issue. The main things that they mentioned were that the communication from CBS was not always as good as it could have been and that they were trying to deal with very complex issues and did not always feel that they got the support that they needed. In particular, the Balquhidder community felt that their procurement exercises were hampered a wee bit by the CBS requirements on who could bid for contracts.
What do you mean when you say that communication was poor? Do you mean that CBS failed to communicate to communities the options that were available and how to access funding and expertise? My understanding was that CBS was more of a clearing point to provide information and bring parties together in order to establish a system.
There was quite a bit of uncertainty around where the two BT contracts and R100 would go, and CBS was not always able to communicate to communities whether they would be covered by the individual projects that were coming up, so communities felt that they were in the dark and unable to make plans because they were not getting the communication that they needed.
I know that there was confusion from BT, because it would not say what it had designated particular areas for. There were grey areas and all sorts of things. Did that contribute to CBS’s failure to deliver? Was the main problem that CBS simply could not identify the areas in which it could support communities?
That was definitely one of the issues. As we point out in exhibit 4, a number of factors came into play, and the uncertainty was one of them.
The report refers to skill levels. The main thing here was finance—it was about financial advice and procurement advice. Are you saying that the people who were recruited into CBS did not have that expertise? Who recruited them?
That is certainly one of the issues. One interesting issue is where the money went, which is set out in paragraph 22. CBS had £7.5 million initially and it spent £6.4 million. Only £2.3 million of that was spent directly on projects, although we could argue that that was because, as you say, its role was to provide advice. As I said, everyone underestimated the complexity and technical nature of some of the advice that was required. Some of the people who were working in CBS did not have the expertise in sufficient quantities or in the right way. To be fair, that is why it drew on consultancy services—it recognised the issue and brought in other people. However, the combination of factors that we set out in exhibit 4 meant that, overall, CBS just did not get the traction that was expected. That is why we say that it is really important that the Government and Highlands and Islands Enterprise learn the lessons of that experience as we go into the R100 programme.
Am I correct in thinking that CBS is now pretty much wound down, except for care and maintenance of the projects that it actually managed to get under way?
That is kind of correct. Paragraph 26 sets out what has happened in that regard. As you say, CBS is much reduced. It does not exist in its old form and now sits within the digital communities team in Highlands and Islands Enterprise. That arrangement is still in its relatively early stages and it is one of the things that we will continue to monitor as the contract for R100 is let in the new year and as we go ahead with it. In a sense, the things that CBS was set up to try to help with will become even more important once we get into the really difficult and hard-to-reach parts of the country. In other parts of the report, we talk about what the Government thinks the £600 million additional investment will deliver. We think that there is the chance that it might need more investment beyond that—the so-called aligned interventions—and at that point the expertise that is available for local communities will be really important.
Are there any lessons to learn from the way that CBS was handled, or was there perhaps an element of inevitability that, in such a complex project, we would come across the difficulties that we have seen?
I guess it is both of those. There are absolutely lessons to learn about ensuring that the right skills are in place, that there is proper resourcing and, for me, that there is clarity of purpose. The communication with communities that Ashleigh Madjitey talked about is absolutely critical, and it will become even more important. That issue was partly about CBS but, as you said, it was also about a need to be clearer on the roll-out of broadband more broadly and who was going to get what and when. The Auditor General made a recommendation on that in previous reports, and we think that it is important to continue with that. There were also some things that were outwith CBS’s control, such as state aid rules and procurement complexities, although we could argue that we now know more about that, so we should be better prepared for it next time.
Colin Beattie characterised the first section of the report on the existing programme as being pretty positive—after all, the 95 per cent target was achieved—but I think that many of us will have heard from our constituents a certain amount of dissatisfaction: they see that the targets have been met, but they feel that they are not getting the benefit of it. I do not want to make this too personal, but I am a case in point. I am connected to a fibre-enabled cabinet, but it is 2km away from me. That means that there are 2km of copper between me and the cabinet, so it does not matter which package I buy—I cannot get remotely the kind of speeds that are being talked out. Nevertheless, I would be counted as a success as far as the programme was concerned. Is that the kind of ambiguity in communication that you refer to in paragraphs 19 and 20?
Lots of people will recognise that situation. The team can help me with this, but I think that this is where the definitions become quite important. The 95 per cent figure that you have mentioned is for access to fibre, but we also refer in the report to about 90 per cent having access to what you might describe as fast broadband of 24Mbps and above. Moreover, as we point out in the report—and as your experience shows—there are lots of variables, not least of which is how far you are away from a cabinet.
When you talk about the 75 per cent figure for rural areas, do you mean that only 75 per cent of properties in those areas can achieve speeds of 24Mbps?
Graeme Greenhill (Audit Scotland)
Exhibit 1 on page 9 indicates the kind of fibre coverage being provided to each local authority area. The 75 per cent figure is for access to fibre rather than speed. As you will see from that exhibit, fibre coverage in the vast majority of council areas is actually more than 75 per cent through a combination of the contracts and private sector intervention.09:15
Yes, but that is fibre coverage. What matters to individuals is the broadband speed that they can get in their property.
That is correct. As the report says, the Scottish Government has redefined what it means by superfast, and it is now saying that 30Mbps is the generally accepted definition of superfast. I think that paragraph 7 on page 8 of the report indicates the number of properties that we are talking about—it works out at around 2.7 million premises.
As Fraser McKinlay has said, exhibit 7 on page 21 shows that there are 376,000 properties that the Government considers cannot achieve 30Mbps at the moment, which means that roughly 85 per cent of premises are capable of achieving 30Mbps. How does that compare with the rest of Europe? Although it does not give separate figures for Scotland, a report produced by the European Court of Auditors gives a figure for the United Kingdom of 85 per cent.
That is great, Mr Greenhill, but I wonder whether the report has fallen into the same trap as the Government programme. When you say that 300,000 properties cannot achieve 30Mbps, does it mean that they cannot achieve such speeds at all or that they cannot achieve them at their cabinet?
It means at their cabinet, does it not?
I stand to be corrected, but, as I understand it, it means that they cannot achieve it at all.
Morag Campsie (Audit Scotland)
As exhibit 7 shows, some of these properties can achieve speeds of 24Mbps.
Are we not missing the point here? I live in a house that is connected to a fibre-enabled cabinet. Because that cabinet is capable of achieving 24Mbps, I am, according to the Government programme, capable of achieving that superfast speed. However, I am not, because I have 2km of copper between my house and the cabinet. R100 could improve that cabinet so that it is capable of providing speeds of 30Mbps, and according to the figures that I think the report is talking about, I would then able to achieve 30Mbps. However, I am not—I am still on 4Mbps.
I share your pain, Mr Gray, because I have been there, too. To be fair, this is where things get quite difficult. The distance from the cabinet is one factor, but there are actually lots of other factors that will affect broadband speeds in your home, and the report will never be able to cover, say, the state of wiring in your house. We have tried to capture that sort of thing in exhibits 5 and 6 in the report; we have used the best data that we think there is, which is on the thinkbroadband website. However, it has its own issues with regard to data, because the people who are more likely to use the website for test purposes are those experiencing poorer broadband speeds. However, that is where the difference is.
What we are trying to report on is what the contracts were said to deliver, which was access to fibre, and then in other bits of the report we have tried to separate out what that means for the speeds that people actually receive. For what it is worth, I absolutely agree that one of the challenges in all of this—and something that I think is a massive issue for the R100 programme—is the expectations gap. As this conversation has demonstrated, this is a very technical matter, but the thing that matters to people is, as you have said, the speeds that they get in their homes. That is the challenge.
My question for Audit Scotland is: if you are auditing performance, would it not be clearer to audit performance on the basis of the speed that is being provided at a property? Obviously, whether I actually get those speeds will depend on the package that I buy, but given that, as things stand, it is simply impossible to get 24Mbps—or in the future 30Mbps—at my house no matter how good the wiring is or how expensive my package is, is it not more helpful for Audit Scotland to audit performance on the basis of the speed that is actually available rather than a hypothetical speed?
We have tried to do both, Mr Gray. It is reasonable for us to report on the basis of delivery against the contract, which is what the first bit of the report—and the initial key message—sets out. In the second half of part 1 and some of the examples that we have highlighted, we have tried to do the best we can to get at exactly the point that you have made.
Finally, the Government and those involved in R100 say that the objective is 100 per cent availability of 30Mbps and above, but that is not actually true with regard to the speeds that 100 per cent of properties can access.
There is always a gap between what is said to be available via the infrastructure and what people will receive in their own homes, and that continues to be the case.
So every house, business and property in Scotland will not have access to those speeds by 2021.
There are, I guess, two things to say about that. First, we do not know the answer to that, because the contract is just being let. The Government is investing a significant amount of money—£600 million—and at the moment we do not know exactly how that will be delivered for the small number of houses that we are talking about. We are going to be in a much better position to answer that question, and that is why as part of our recommendations we have said that, by next summer, we expect to see a very clear plan.
That is what I am asking. The target is not that every single property, house and business will have access to those speeds.
This is where we slightly get into semantics. I do not think that this is a target that says, “Every house in the land will have 30Mbps”, because, to be fair to the Government, there are lots of variables that are genuinely outwith its control. That is my point about the risk of an expectations gap between what the consumer and end user—me, you and everyone else—is looking for and what the infrastructure is actually going to provide.
I do not understand why you cannot audit the performance on the basis of the question that matters to people: if I buy a superfast broadband package for my house, will I get superfast speeds? Would it not be a lot simpler to audit performance in that way?
I think that it would be an awful lot more complicated.
Because it is quite hard for us as auditors to know exactly what speeds people are getting in their homes.
I am talking not about what they are getting, but about what is actually possible.
That is what I mean.
Obviously, the package that you buy has an influence—
Alongside a lot of other variables.
I am sorry, but which other variables?
There are variables such as the distance to the cabinet, the wiring in a house, the number of people using the internet at any given time and so on. In the report, we list a whole range of different variables that will affect the speed that you get.
I think that that is an illustration of why there is such dissatisfaction. People hear that the programme has succeeded, but they still feel that they have been let down.
I think that Iain Gray speaks on behalf of frustrated consumers everywhere. Mr Neil has a supplementary.
I have had similar experiences with constituents of mine, particularly in more rural areas. As the infrastructure secretary who issued the initial contract to BT—
—I have some experience in this matter, and I cannot see why BT, which is the supplier, cannot tell you the number of people each cabinet can provide with superfast broadband and the speeds that they can get. It knows these things: it is its copper that runs from the end of the fibre to the house. You have the cabinet; you have the fibre; in Iain Gray’s case, which is a very good example, you then have the copper and the wiring in the house. The question is not how many houses have or use it, because that is a decision for each house, but BT can tell you how many people serviced by a particular cabinet get the fibre straight to their house and how many do not. That would at least indicate by proxy the people who are genuinely going to get superfast broadband and the people who will have much more difficulty getting it. Certainly—and this might be a presentational point—the impression that everybody has—
Can you ask a question, please, Mr Neil?
I am just asking whether this is correct. The impression that everybody has is that the target is 30Mbps or 24Mbps for whatever percentage, but it is not qualified in the way that you have suggested, Mr McKinlay. In any case, BT should be able to give you that information, and I do not see why you cannot ask for it.
I will ask Ashleigh Madjitey to come in on the specifics of that but on your specific point, BT would be able to tell us how many premises have fibre going into the house, because fibre to the premises, which gives the fastest available broadband, is actually pretty unusual. However, that is a different thing from the point that I think Mr Gray is highlighting, which is about the speeds that those who do not get the fibre and are connected instead by copper actually achieve. I will ask Ashleigh Madjitey to provide some detail about that.
In exhibit 2, we have set out the modelled speeds that BT has said will be received by people who have been connected through the contracts. In the contracts, it committed to delivering speeds of 24Mbps to 77 per cent of everybody connected to the fibre network, but it has turned out that 90 per cent of everybody connected to the network are receiving speeds of 24Mbps.
That is the assessment of how many people should be able to get 24Mbps or more, taking into account the factors that we have talked about.
Maybe a recommendation should be that the Scottish Government carries out some pilot testing or surveying to find out how accurate the figures are. I do not think that we should just accept BT’s figures as God given, because that has not been our experience.
In exhibits 5 and 6, we have tried to reflect on the thinkbroadband data, which is based on exactly that: people’s real and live-tested speeds. There are caveats associated with that data, but it is the best available for telling us the speeds that people are actually experiencing, and it is where you can see a gap.
I think that Ashleigh Madjitey also wanted to refer to exhibits 5 and 6 to underline her point.
Yes. I wanted to underline the gap between what BT has modelled, as shown in exhibit 2, and what people experience when they test their speeds on the thinkbroadband website. This is about as accurate as we can be about the speeds that people in Scotland are actually experiencing—with the caveat that people are more likely to test their speed if they are unhappy with it. That might be an issue here.
Do members have any further questions on this point? Mr Coffey, is your question about this point?
I have a number of points that I would like to ask about, convener.
Please ask your question on this point. I will then take Liam Kerr and come back to you afterwards.
There is no doubt that the remaining 5 per cent or so is going to be the most difficult to achieve, although the commitment has been made. As Fraser McKinlay has said, the contracts will be negotiated and agreed in early spring 2019.
As for Iain Gray’s question about his cabinet, there will be people attached to it who will get the required speeds, but, because he is further away, he will not. Of course, that is a challenge, and I certainly expect to see the Scottish Government and its contractual partners planning to solve that, because it is the biggest challenge that they face.
However, I do not want to sound as though the project is covered in gloom. This is a fantastic investment; indeed, had it not been for this investment, it would have taken us years and years to get anywhere near this kind of standard. The European Union standard of 30Mbps is a good one to embrace, I think.
Can you ask a question please, Mr Coffey?
Well, I just want confirmation of some of the things that have been said here. The procurement for the remaining 5 per cent is taking place now, with the contracts awarded in spring 2019. Perhaps things are not quite at the stage that Iain Gray is looking for, but when will we see details of how the hardest-to-reach customers in Scotland will be reached?
As you have said, the tender is out just now, and we are expecting the contract to be let in early 2019. That is why we are saying that we think the Government and Highlands and Islands Enterprise should be in a good position to have a clear plan by next summer about exactly what their investment of £600 million is going to deliver and potentially what will be left over and what will remain to be done. In exhibit 7 on page 21, which has already been mentioned, we have tried to set out as best we can what the plan is to reach the 376,000 premises that currently do not have superfast broadband.
I would like to explore that point, which was well made by Willie Coffey. We have the R100 programme, for which the stated ambition is that by 2021
“every home and business in Scotland will have access to”
superfast broadband. I understand Iain Gray’s frustration with that. You said, “To be fair to the Scottish Government”. I am not sure that we can be fair to the Scottish Government. That is a cast-iron commitment: every home will have that. I think that the point that Iain Gray is making is that that is not true, is it not?
Again, we come back to the question of the difference between “have access to” and what people experience in their own home. I know that for most—
I will come back to that. I have the expectation, because of the commitment that has been made—as I think that Iain Gray is pointing out—that that is what I am going to get, but that is not the reality. Is that not what you are concluding?
We are concluding that that will be difficult to achieve. I do not think that we are at a point now to conclude one way or the other for sure, because the £600 million that is being invested in the R100 programme is out to tender and we need to wait and see what that will deliver. Based on the current plans, which we have set out in exhibit 7, we absolutely think that that will be difficult to achieve.09:30
What does “difficult to achieve” mean, in the speak of the people watching this meeting? Will it happen? Will there be superfast broadband in every home and business by 2021?
I cannot say yes or no to that, I am afraid, much as I would love to be able to. The Government has set out significant additional investment of £600 million. We wait to see what that will deliver and the extent to which that will deliver the 100 per cent. The Government has already said that there is the aligned interventions programme to come, which might include a voucher scheme and various other things, so we need to wait and see what that will deliver.
As we have set out in paragraph 49, the reasons why we think that it will be difficult are, first, that with the contracts being awarded in early 2019, that leaves less than three years to put the infrastructure in place for the most difficult-to-reach premises, which are the only ones left. Secondly, the aligned interventions programme is under development just now, so work is under way, but a lot will depend on what comes out of the R100 procurement before it can be decided what is needed in terms of aligned interventions. There is still quite a lot of work to do and a lot of work in progress. It is quite a short timescale and, as I said, the most difficult-to-reach premises remain.
What does the tender invitation for the provision of R100 say? What is the specification that any successful tenderer will have to deliver by 2021?
As we set out in paragraph 40, the tender is split up into a number of lots. The Government has also set out some mandated areas and some weighted areas. As we have pointed out, the aim is to deliver 30 Mbps. There is obviously fibre to premises and fibre to cabinet, but the Government cannot set out in the tender a percentage that it would like in terms of how many connections are going to premises and how many to cabinets. In paragraph 41 we point out the mandated areas and the weighting that the Government set out in the tender. The hope is that that will encourage the bidders to try, as far as they can, to provide a full-fibre solution.
As we say in the report, further down that page in paragraph 45, it looks unlikely to us that the £600 million will deliver the R100 programme, which is why we say that additional investment may well be required. In a sense—
Forgive me, but I really want to explore that £600 million. I am slightly concerned by Morag Campsie’s response there, not because it is your response, but because you used the words “the hope is” that that will happen and that it will “encourage” the tenderers. What I am hearing is rather different from, “We will deliver this by 2021,” although I appreciate that you are just reporting back. The Scottish Government has said that £600 million of public money will make this happen and has allocated that money. Audit Scotland is in a position to say that, even on the very basic spec that we are hoping for, or that the Scottish Government is hoping for, £600 million is unlikely to be enough. Why is Audit Scotland able to say that, whereas the Scottish Government, which is responsible for administering that public money, has apparently not done a similar level of analysis? Is that not concerning?
I should probably stop saying, “To be fair to the Government,” but the first thing to say is that I do not think that the Government has said that the £600 million will deliver the R100 programme. It has said that that is an initial investment, which, as we have tried to set out in exhibit 7, will get it a good way along the road.
We are challenging the Government. I have talked about it delivering a plan next year, once the procurement process is under way. The Government is already, even before the 100 per cent is reached, excluding 30,000 urban premises and 34,000 premises that can already get 24 Mbps. It is not clear yet what the plan is to plug that gap and it is also not clear, as you can see down at the bottom of exhibit 7, how many of the 147,000 premises that are within scope will actually get 30 Mbps.
The Government has committed the £600 million. It is waiting to see what the procurement that Morag Campsie talked about and that is described in paragraphs 40 and 41 will deliver. We think that there will then be a good chance that it will need to think about how it reaches the further bits. I think that it is absolutely clear to most people that a target of 100 per cent of anything is always a difficult ask. Equally, we are not in a position, given that the contract has not been let yet, to say that it either definitely will or it definitely will not happen. We need to wait and see.
It is not a target; it is a commitment. My concern is that £600 million will be put in and you are telling me that at some time in the next two or three years somebody will come back and say, “I need a significant further investment to deliver that commitment.” Either the Government will say, “Fine, here is some other cash that we will pull from somewhere” or, presumably, it will say, “There is no further cash. The commitment will not be delivered,” and Iain Gray will be sitting here in 2021 saying, “Superfast broadband in my house still does not work.” Is that not the case?
That is possible. Once it knows what the initial £600 million will deliver, the Government will need to decide what it does next, which is why we are absolutely clear that we need a pretty concrete plan in the summer of next year, once we know what that money will deliver.
It should be clear what they have to do. The only solution for me—not to focus too much on me—is to replace that 2km of copper with fibre. Nothing else will make that difference, so unless the contract is going to do that, then—
At the end of the report, we set out other ways of delivery.
I will not qualify for those. It is correct that superfast broadband could be delivered by microwave or wireless, but I will not qualify for those because I am already connected to a fibre-enabled cabinet, so I will not get the voucher.
No voucher scheme has been agreed yet and that, again, is where we need to see what the plan is next year, once we know what the £600 million will deliver. As we get to the end of the process—this is a point that we make in the report, I think—there are some really important questions about value for money. This is another £600 million of investment to get towards the 100 per cent. As we get to the really remote and hard-to-reach places, some of the costs could be quite significant, particularly if you are talking about putting fibre right into the premises. Some of the numbers are up in the thousands, if not tens of thousands of pounds per premises, so there will be a question at some point about a balance between what people can get and how much it will cost.
Colin Beattie rightly looked at the people who were in place for the CBS project earlier on and it would appear that perhaps the skills and the right people were not in place for that project. Are the right people in place to deliver the R100 project? If so, who are they and how confident can we be that we will not be sitting here in three years’ time looking at the leadership and the skills that were in place and having the exact conversation about R100 that we have just had about CBS?
That is a key question for us and we recommend exactly that, in that the lessons must be learned, there must be a degree of consistency in the team and the right skills must be in place.
There are people on the R100 team who have been involved with the digital Scotland superfast broadband programme. The community broadband Scotland team has also been working with the R100 team, who have been trying to learn lessons from that. Some parts of the team—I mentioned the aligned interventions team—have just been put in place over the summer, so they will be building up experience. As Fraser McKinlay said, our report recommends making sure, in particular around contract management, that they have the right skills and finance management in place. The Scottish Government has been building up that team over the course of the last six to nine months, and we will be keeping an eye on that.
Paragraph 25 in the Audit Scotland report clarifies that a commitment has been made. Colleagues are right to question the commitment, which is that every home, including Iain Gray’s, will have access to superfast broadband by the end of 2021. Superfast cannot possibly be what you get, Iain. Is it 2Mbps?
It is 3Mbps.
The commitment is there, but the proof is in what happens during the early phase in 2019 and how the programme will be rolled out. There are a number of ways of trying to deliver superfast broadband to premises and homes such as Iain Gray’s. That is to be worked out, but the commitment is quite clear: superfast will be delivered to every home by the end of 2021. I am looking forward to seeing how that will be possible in some of the remote and rural areas, but the commitment is definitely there.
Do you have a question for the panel, Mr Coffey?
Can I nip back to community broadband Scotland?
Did the community broadband Scotland team not have access to all of digital Scotland’s skills and expertise? It is as though it operated completely in isolation. Is that perhaps what Colin Beattie was talking about regarding skill levels? You mentioned that in your report. Was that an issue in community broadband Scotland? Why did it not have access to the same skills base, which could have provided that advice so that a lot of money did not have to be spent on going down the legal route?
What happens on a continuity basis when technology inevitably moves on? Will community broadband projects be embraced as new technology emerges?
The team had access to those skills and it worked with colleagues. As I said, the skills issue was one part of the CBS story, but, to be fair, there were many others. What is important is that people are now trying to embed those skills more centrally.
Your point about technology moving on is an important one. Part of the discussion that we have just been having is that this technology is moving very quickly. It may be that, in even one or two years’ time, new technologies and approaches come on stream that make some of this stuff more doable than it is at the moment. Certainly, the approach that the Government and HIE—and whoever they end up partnering with—are taking to R100 will be taking account of that fast-changing technology environment.
In your thinkbroadband exposure in the report—and you give it a lot of exposure—it was good to see Dundee at one end of one chart and the other end of another. However, I know that constituents in Dundee will tell me that they would love a 10th of the speed that is presented in exhibit 5. I do not know whether you talk about hotspots or coldspots, but what credence do you give to those figures? You have put them out here. Do you just accept them?
No, we recognise that the picture is far from perfect.
I do not see a health warning on the front of the report.
There is a bit of a health warning in the report. The note for exhibit 5, for example, tries to explain some of the issues with the thinkbroadband data. In a sense, this is the best data that we have available for the speeds that people actually experience. If anything, we think that the figures are likely to be pessimistic rather than optimistic because, as Ashleigh Madjitey said, people are more likely to test via the thinkbroadband website if they are dissatisfied with their broadband speed.
Some of the figures are on download speeds.
Yes, but thinkbroadband also tests for uploads. We just have not presented those figures.
I suspect that people in places that do not get good speeds have just given up on testing. I would have hoped that there would have been a little bit more of a warning in the report than—
We also compared the figures to Ofcom’s data, but it is gathered less frequently. Ofcom’s data was from 2017, so it was a bit out of date, but it was telling the same story. The order of councils in its results was very similar. It backed up our data from thinkbroadband.
If I was a broadband company, I would be very pleased that you are publishing how wonderful everything is.
Members have no further questions for our witnesses, although we probably have a few follow-up questions for the Government. I thank Audit Scotland for its evidence this morning.
I suspend the committee for two minutes to allow a changeover of witnesses.09:45 Meeting suspended.
09:48 On resuming—
Section 22 Report
“The 2017/18 audit of the Scottish Public Pensions Agency: Management of PS Pensions project”
Item 3 is the 2017-18 audit of the Scottish Public Pensions Agency. I welcome our witnesses from Audit Scotland: Caroline Gardner, Auditor General for Scotland; Stephen Boyle, assistant director; and Tom Reid, senior audit manager. I invite the Auditor General to make a brief opening statement, in which I believe she will cover the legal status of the report.
Caroline Gardner (Auditor General for Scotland)
Thank you, convener. The purpose of this report is to make Parliament aware of problems with a major new information technology system for the Scottish Public Pensions Agency and the financial implications of the delays. The agency’s main role is to administer and pay pensions for members, deferred members and pensioners of pension schemes of the national health service, teachers, police and firefighters in Scotland.
In October 2015, the agency awarded a £5.6 million contract to Capita Employee Solutions to deliver a bespoke unified pensions administration and payment system. The new system, PS pensions, was to be operational by March 2017. In February 2018, the agency’s chief executive decided to close the project. The agency has spent £6.3 million on the project and has written off £1.6 million in capitalised assets that will no longer be used. The closure of the project means that the agency has not been able to progress its strategic, business and workforce plans as originally intended. It has forecast significant budget gaps and estimates that it requires an additional revenue budget of £9.8 million between 2019-20 and 2022-23. It also needs a total capital allocation of £18.4 million to deliver the replacement project.
The agency has extended contracts with its existing suppliers to ensure that payment of pensions is not affected by the closure of the project. The SPPA is in discussions with Capita to establish whether either party bears responsibility and potential financial liability due to the unsuccessful project implementation. I have, therefore, not commented on arrangements for implementation of the PS pensions project. Once the legal process has concluded in 2019, I intend to prepare a further report that will cover those issues.
As always, we are happy to answer the committee’s questions about the current position and the likely financial implications, but we are unable at this stage to answer questions about the causes of the problem. We will come back to that, as I said, in a report next year.
Thank you very much, Auditor General. Alex Neil will open questioning on the report.
I have two questions. The first one is: who picks up the tab? Clearly, irrespective of the outcome of the negotiations or any legal action, there already has been cost and there is likely to be further cost to the pensions agency, in terms of both administration and the capital budget. I am not aware of the relationship between the Scottish Government and the pensions agency. Who provides funding for the pensions agency’s core running costs and capital costs? Is it provided by the Scottish Government or is it self-administered? Does the agency get a fee from the pensions providers or whatever?
You are right: whatever the outcome of the discussions between SPPA and Capita, there will be financial implications, both revenue and capital. Stephen Boyle will talk you through the current discussions about who will meet those costs.
Stephen Boyle (Audit Scotland)
The SPPA is an executive agency of the Scottish Government, so it is funded by the Government to meet both its revenue and capital requirements in the current year. In its forward planning, it captures its requirements and expectations for revenue and capital into the future.
What we have not set out in the report is the implications of the situation and whether there are any financial penalties for either party. If there were any recoveries from one side or the other, they would need to be captured in forward financial plans, particularly of the SPPA, and they could potentially offset any future capital requirements that the SPPA has.
Whatever cost there is would come from the Scottish Government’s budget.
Yes, that is correct.
So, there is a cost to the taxpayer of this fiasco.
Yes. As I said, the SPPA is an executive agency of the Scottish Government and it is funded directly by the Scottish Government.
Obviously, we do not want to get into who is to blame yet, because that is hopefully going to be resolved by the negotiations and/or legal action, but at some stage you will come back to this committee after you produce your 2019 report, Auditor General.
The wider issue is that we were given reassurances by the permanent secretary and others in the past year that IT projects were now under much tougher, better management and that there was better control and all the rest of it. Here we have again—irrespective of who did what and who is to blame and irrespective of the outcome of the negotiations—another IT project that has gone badly wrong. Does that mean that, despite the promises that were made, we are still in a position where Government does not seem to be able to get big IT projects right?
I cannot answer that question at this stage. There are discussions under way between SPPA and Capita Employee Solutions about what has happened with the project itself. I do not want to compromise those discussions, but I will report back next year once they are complete.
All right. Did the project predate the commitments that were given to this committee about the introduction of new and allegedly better management of IT projects by Government and its agencies?
The contract was awarded in 2015 and planning would have started well before that.
So, it does predate them.
Auditor General, I would like to take a little look at the governance of the SPPA. As an executive agency of the Government, I presume that it is subject to the Scottish public finance manual. The project would therefore be a major investment as it is more than £5 million, inclusive of fees and value added tax. Is that correct?
Stephen Boyle, do you know what the position was at the point at which the contract was awarded?
I do not have that in my head, Mr Beattie. I would need to come back to answer your question.
Is the cost of it more than £5 million?
Yes, it is.
As set out in the document that I am looking at, it would therefore fall under major investment.
On 31 May, Colin Cook from the digital directorate wrote to the convener assuring the committee that, since January 2017, the Scottish Government has taken a more interventionist approach. He detailed carefully all the steps that are being taken, all the different processes that would make sure there would be no issues in the future, and that there would be engagement with the corporate assurance providers, internal audit, capability development, project assurance team, and so on. Is there any indication that that has taken place in the project?
Again, I am afraid I have to say that I do not want to comment on the management of the project at this stage because I do not want to compromise the discussions that are under way between SPPA and Capita Employment Solutions. I know that is frustrating for the committee, but the purpose of the report is to make you aware of the issue and of the potential financial implications rather than focusing on where the responsibility for the problems lies.
I appreciate that you have difficulty in commenting on some of these things, but perhaps I can just mention one other thing about the communication of 31 May. It lists all the IT contracts, including the one for the SPPA. You would have to have a considerable degree of perspicacity to work out that, where it says “closure” and, at the same time, that the project delivery date is February 2018, you might infer that that was actually the project delivered. You would then have to work out that where it says “business justification” and so on underneath, “The preparation of variations of existing contract”—
Mr Beattie, I am sorry to interrupt, but the Auditor General has been quite clear on what she can and cannot answer here. You seem to be going down the line of fault. The Auditor General has been clear that we will come back to that next year, so if you are asking about the figures, I will allow you to proceed, but if not, I will move on.
Clearly there is a fair bit of restriction on what we can discuss at this point, which is unfortunate.
That is fair. It is unfortunate, but I think that the Auditor General has been clear and it is for good reason. She will come back to us next year.
In that case, I will leave it at that.
I have three questions, and I will try to keep two of them general so that we do not get drawn into the limitations that you have set, convener.
Alex Neil has referred to some of these so-called fiascos and you probably know, Auditor General, that the projects that have not worked well all come under the public sector umbrella. If staff members are involved in those projects, does their involvement travel with them if they move jobs? If someone in another agency is setting up a project, would they know that somebody has been associated with a project? That was one question.
Secondly, given the huge market for public sector and large organisational systems, why are bespoke solutions required? Are there no off-the-shelf solutions that would be good enough?
Finally, you mentioned that the original contract was for £5.6 million, but £6.3 million has been spent and £1.6 million written off. Where does the £1.6 million come in? Is the write-off not actually more? Could you go through those questions, please?
I will do my best. I will ask Stephen Boyle and Tom Reid to answer the third one for you in terms of the accounting treatment.
The point about the skills question is generally the point that Mr Beattie was making. The Government has recognised that a number of significant IT systems have not delivered as planned or in some cases at all, and it is difficult for public bodies to build up the skills and the expertise that they need to be intelligent purchasers of IT systems or to develop them themselves when they decide to take that route. The Government has been investing in a central information offices team who can provide that support and challenge to projects.
That is one of the things that we will be looking at when we report back next year, once the legal discussions have concluded, but it is made more difficult by the fact that IT skills are in short supply and tend to attract a premium in terms of salary. We will be looking at how well that has worked in practice for the project as well in general through a piece of work we are planning to do on digital transformation in central Government.10:00
Could somebody move from one problem project to a new job and do the same thing?
The aim is to do exactly the opposite, and take people who have moved from a successful project—and there are successful projects; they do not tend to come to this committee, but they do exist—and to use those skills and expertise on other projects.
You do not think there is a flagging system or some way of knowing that somebody has been associated with such a project.
I do not think that there is necessarily a flagging system, but the Government takes very seriously the need to improve its performance in these areas and is making sure that people who have been responsible for a problem system do not move on to another one. The aim is to do exactly the opposite, and build capacity and expertise. I cannot comment on what has happened with the project, but that is certainly the intention.
On bespoke systems, sometimes there are bespoke systems, sometimes there are not. We expect to see that in the planning and the business plan option appraisal for any particular system. Again, I cannot comment on this one at this stage. There has often been reluctance to look at whether you can get 90 per cent of what you want from a commercially available system rather than running the risk of the likely extra costs of a bespoke one. The central information office is trying to develop expertise around that.
Is it actually looking at that particular aspect?
For the SPPA project?
No, I mean generally.
In general, yes. I think looking at the make or buy question is central to what it is trying to do and will be part of what it will do in the planned piece of work on digital and central Government next year.
Stephen Boyle or Tom Reid can pick up on your third question about the £6.3 million and £1.6 million.
Tom Reid (Audit Scotland)
As we say in paragraph 9 of the report, the £6.3 million covers all the costs associated with the project, including the costs and the money paid to Capita as well as staff costs, hardware and other elements.
In terms of the £1.6 million over the development of the contract, the hardware and elements of the software development were capitalised and recorded in the annual accounts. The £1.6 million represents what was written off in the agency’s annual accounts.
Is that in addition to the £6.3 million?
It is part of the £6.3 million. We just highlight it separately because, as I said, that is an amount that has been written off within the agency’s accounts.
You can spend whatever £6.3 million minus £1.6 million is and it brings no value to the business and it is not capitalised. That surprises me slightly.
No, the £1.6 million has been capitalised. It is also to do with the timing of the closure of the project. Costs were incurred within the year that would not have reached the point at which they were capitalised.
If the project had been going ahead, that would have been capitalised.
Elements of it would have been capitalised.
I know that there are accounting rules, but if you are spending that amount of money on what does not seem to be going to be an asset, I have to wonder what the project is actually all about.
Auditor General, can you clarify that?
Stephen Boyle will be able to do that.
We can expand on that more fully in the report that we plan to bring back to the committee next year.
Convener, I think that it is probably wise not to investigate the project any further and put the Auditor General in a difficult position. I would like to ask a number of questions about software procurement and methodology, but they can wait.
Can they wait until next year’s report?
I think so.
That is very good of you, Mr Coffey, thank you. Liam Kerr, do you feel the same or do you have a question?
Willie Coffey is right. It would be interesting to explore an awful lot of this.
I have one question that I will try to phrase correctly. It might be interesting looking forward. The original contract was £5.6 million. Tom Reid has just talked about there being a spend of £6.3 million, but the total capital allocation required in the SPPA estimates going forward for five years will be £18.4 million, which is a considerable difference from the original estimate. Do you have any information on why the cost of delivering the replacement system—the £18.4 million system—is going to be that much higher?
That is the complete capital budget for the agency over that period of time, so it does not just reflect the costs of the project itself. It covers the extension to the incumbent supplier, and the agency is budgeting £6 million for that element. It represents the costs of a new procurement and implementation. The agreement that the agency has entered into with the incumbent supplier goes to 2024. They will need to look at a longer-term solution going forward, so an element of the budget covers the longer-term solution. The final element is miscellaneous capital, which covers areas such as existing information and communication technology and existing capital works on the agency’s headquarters.
I see. I just want to check that I have got this. A significant element of the £18 million is what Alex Neil was exploring—the run-on of where we are at the moment, just to make sure that everything stays in place and nothing goes wrong in the meantime. Can you isolate how much of that £18 million is the cost of a new package, such that we can say, “Okay, it is comparable to the £5.6 million that was originally projected” or not?
There are two elements—or three, in effect—to the approach that the agency has taken. It has extended the contracts with their existing suppliers to ensure that it can continue to pay pensions. It has entered into a contract extension with one of the incumbent suppliers and, as part of that agreement, it is looking to develop the system so that it can start to develop a unified, bespoke system. That takes the agency to 2024. It is aware that, beyond that, it will have to look at a more long-term solution and this gives it time to do that. It is currently budgeting £7.4 million for a new procurement and implementation of a further system. As I said, that is just what it has budgeted at the moment. It is a longer-term approach, so the costs are not as certain as they will come to be.
I understand that. Again, just to be clear, about £7.4 million is the comparable figure to the £5.6 million that was budgeted for back in 2012, or something like that?
Thank you very much. We look forward to coming back to this, Auditor General, with some more questions once you report again in 2019. I thank you and your team for your evidence this morning. I now close the public part of today’s meeting. Thank you.10:07 Meeting continued in private until 10:30.