Chamber and committees

From looking at the detail that has been presented and from hearing some of the answers today, one of my key concerns is the level of granularity at which the project has been monitored and managed. Mr Carlaw said that it is not possible to specify which bit of spend would have occurred in which year, but I suggest that that is precisely what you should be able to do in a well-run IT project. In a well-run IT project, as I said before, you should understand what levels of effort were required at each phase, and have a project management office that tracks those things, using things such as Gantt charts and projected spend. That seems not to be what we see. I would be grateful if that detail could be provided.

In particular, I am very concerned by the category “non-technical Contractors”, which accounts for almost a third of the total costs. That is a very non-specific category of work. Can you explain why you were using third-party contractors to deliver the project? Managing individual contractors on an individual basis seems to be a very complicated way of managing a project like this, but it seems to be what happened.

Dr French, could you pick up on that and also say whether there is a sense in which the metrics need to be split apart from the capturing of the outcomes? I accept what Dr Elliott is saying, but I think that, if we just had qualitative outcomes with no measurement, we would have a problem. At the focus group that I attended in Glasgow—there were parallel focus groups in Dundee and Glasgow—there was a view that we are not using data properly and that we have narrow metrics, which is a problem because, in the 21st-century world, people use big data sets and do much richer data analysis.

Do we need to split apart the capturing of the outcomes from the measurement, and do we need to overhaul how we conceive of what the measurement looks like so that we can capture that 21st-century big-data approach?

Okay. I could ask more questions, but I do not want to use up more time.

I would be grateful if somebody could write to the committee to clarify a couple of things. A number of boards and management groups have been mentioned; it would be good to get some documentation on how they interrelate and how the governance works.

Secondly, if the agile methodology was used, I would be grateful for an explanation as to why, because it strikes me that the project had a clear functional footprint, so I wonder whether the agile methodology was at all appropriate for delivering a website project such as this one.

Indeed. However, in terms of the overall budget, it is a significant project. In order to understand whether it was getting out of control, we need that level of detail.

However, the problem with transparency goes a little further. On the detail that has been provided by the SPCB about how the costs are accounted for, all that we have been provided with is a schedule of resource costs set out according to whether they were for technical and non-technical contractors, but there is no specificity about what work they were doing.

I would expect, in any IT project, to see phases split up, so that we could understand where efforts are being applied—whether to initial analysis or to the design, build, testing or user acceptance phases, for example. Those are very basic things, but we do not have that level of detail. Why has it not been provided?

I do not understand why, in table 2 of your written submission to the committee, you have described categories as being “areas of spend”, but that is not what they are. The table is a time profile that is broken down by resource type. In order to get a good handle on any IT project, you need to understand that effort by phase, do you not?

I like that phrase “radical decluttering”. It sounds like what I am constantly being told to do at home.

I am very interested in the conversation so far. I will pick up on a couple of points.

Structure seems to be the thread that runs through the conversation at a number of different levels. Dr Ian Elliott talked about the national performance framework originally being a decision-making tool. The outcomes are relatively straightforward to understand. However, when you come to the indicators, you are suddenly landed with a sea of bullet points and it is difficult to see intuitively what they are trying to tell you or even whether it is one thing or a number of different things.

I looked at the indicator on children’s happiness. It turns out that that is just one survey that manages four quite narrow metrics, which are valuable but do not necessarily entirely encompass what we would all understand to be children’s happiness.

Is one of the problems the structure of the NPF, in that we have good high-level outcomes with an asymmetrical set of indicators that sit below them and it is not intuitively easy to understand what any of them tells us? In other words, is the NPF as it is currently structured too difficult to use?

I ask for a clarification. You talked about making things more accessible. There also seems to be a point about interpretation. Are you saying that, both at Government-wide level and individual directorate or agency level, there needs to be an interpretation of how outcomes are going to be influenced? It strikes me that two people will have completely different ideas on what would impact on outcomes and, unless there is a stated view in that regard, there will be no consensus. Is that a fair interpretation of what you said?

Leading directly on from that, and similar to my previous line of questioning, it strikes me that, over the past 10 years, we have had the credit crunch, Brexit, Covid and, now, the war in Ukraine; it seems that our black swan events are turning into a bit of a flock.

Having looked through the Fiscal Commission work to date, I see that it has responded to those things. However, I do not necessarily see, either in the body of its main forecasts or in what it publishes more generally, a risk register, for example, or a forward look at contingencies and potential risks. Are those things that should be thought about in terms of some sort of counterfactual assessment and longer-term forecasting?

You have highlighted the specific case, so, before moving on, I will characterise what is in the Audit Scotland report. From paragraph 20 or so onwards, the report shows that a preferred bidder status was awarded on the basis that the ferries contract would be a standard contract in which the constructer assumed the risk. The contract was then revised so that a 25 per cent risk was assumed by, in essence, the public purse. That issue was flagged up, but Scottish ministers still apparently approved the decision. However, there is no documentation of that approval. That is not acceptable, is it? Do you agree that, when a preferred bidder status is awarded on a certain basis and the contract is then altered, that critical ministerial decision should be recorded?

I am almost tempted to leave that as the final word, because it is so important.

Nonetheless, I want to follow up on some of the points that have been made about freedom of information requests. I challenge the point about the importance of legal advice, because I do not believe that the issue is limited to that.

On 8 April, the Financial Times published an article that resulted from a freedom of information request on communications on its original FOI request regarding the Gupta guarantees. Among those communications, there was an email from 29 September between civil servants in the Scottish Government in which the following was stated:

“Here is the long-awaited decision in the Lochaber smelter appeal. Unsurprisingly, the Commissioner has not upheld our s.33(1)(b) arguments, as we have been predicting since at least the review stage. That said, I imagine this is not what Economic Development colleagues were hoping for. I’ll start thinking about what we say to them”.

The point is that it is very clear that officials in the Scottish Government were knowingly withholding information following requests, when they knew that it was highly likely that that decision would be overturned by appeal. Furthermore, those final sentences seem to suggest that there was internal pressure on them to do so.

It is one thing to withhold information on principle, and another to defend that on request. However, when you start knowingly to withhold information, while knowing that you are highly likely to have to reveal that information on appeal, are you not into slightly different territory? Are you not actually knowingly withholding information, and is that not suppression?