Reinventing a faulty wheel

In an extensive piece of research, Margaret and Jim Cuthbert discover not only that many of the criticisms of PFI have turned out to be correct but that the mistakes are being repeated by the current Scottish Government

This paper reports on a study we have carried out analysing the bidding process for schools PFI projects in Scotland. This tells us a good deal about PFI, and about some important things which have gone wrong. Unfortunately, even though the current Scottish government is attempting to correct some of the worst features of PFI, we will see that it has failed to learn key lessons.

In Scotland, there have been 37 schools PFI projects to date: the first project, Falkirk, was advertised in 1998, and the last contract was signed in 2009. Overall, the 37 contracts involved around 245 new schools and 45 refurbishments and/or extensions. The capital value of these projects was £3.3 billion – although, of course the total that local authorities will pay over the lifetime of the projects in unitary charges, (covering not only the capital financing costs, but payment for the provision of services), will be much larger. For example, the total in unitary charges paid by local authorities in 2009-10 is estimated at £390 million: rising to £465 million in 2010-12 when all existing PFI schools projects are in operation.

Our study looked at the bidding process for each of these 37 projects. Our approach was to obtain the Final Business Cases of all schools PPP/PFI projects in Scotland. Some of these are available on individual local authority websites: most, however, were obtained through Freedom of Information requests. The information has been supplemented by minutes of Council meetings, and by contractor bulletins on PFI. General statistics are from Partnerships UK and from the Scottish Government. We examined this information on the bidding process to see what it told us about the degree of competition, or other potential problems. The first thing we looked at was project size. Sheer size is an important factor in PFI for two reasons: first of all, because the number of firms capable of bidding for very large projects is likely to be severely limited – so that very large contracts in themselves are likely to be associated with a restricted degree of competition. And of course, from a Scottish perspective, the larger the contract, the less likely that a Scottish firm will be able to handle it – so very large contracts are likely to be associated with the loss of Scottish market share.

What we found was that the projects did indeed tend to be large. Remember that there have been 37 contracts in all covering around 290 schools. The five largest contracts alone accounted for 107 of these schools. At the other end of the scale, only 14 of the projects were for less than five schools. In terms of capital value, there was one contract with a capital value of over £300 million, two in the range £200m to £300m, and twelve more over £100m. The largest individual contracts were Glasgow with 11 new schools, and18 refurbished; Edinburgh with 15 new and two refurbished; North Lanarkshire with 24 new; South Ayrshire, 17 new; and South Lanarkshire, 17 new and three refurbishments. So, PFI schools contracts in Scotland are indeed large. Moreover, they are relatively large in UK terms. Figures from Partnerships UK indicate that the three Scottish projects are the only ones in the UK with a capital value of over £200 million. And of the 10 UK schemes with a capital value of over £150 million, six were in Scotland. In fact, Scotland has not just large individual schools PFI projects but has a disproportionate large share of the overall market in schools PFI over the UK as a whole: Scotland, with just 8.5 per cent of the UK’s population, has 40 per cent of the UK’s PFI schools projects, as measured by capital value. We will return later to the significance of this.

The next question we looked at was: how competitive was the market for Scottish PFI schools projects? There are a number of ways of looking at this. One approach is to consider the number of firms, or groups of firms, engaged at key points in the PFI bidding process. One key milestone is the issue of an Invitation to Negotiate to a number of consortia. A later milestone is the final selection of a single preferred bidder from the resulting bids. The view taken by the Treasury and the Public Accounts Committee is that if there are two or fewer compliant bids at the final selection stage then this is potentially indicative of a failure of competition, and special care should be taken: indeed the Treasury lays down specific requirements which must be met before single bid projects are proceeded with. In fact, of the 37 Scottish schools PFI projects, two had only one viable bid at the final selection stage, twenty-eight had two bids, and only seven had three bids.

These figures in themselves are suggestive of a fairly restricted degree of competition in this market. Nor is it the case that the small number of bids at the final selection phase arises solely because bidders are dropping out at a late stage in the bid selection process. For ten out of the thirty seven projects, there were only two or less bids at the initial Invitation to Negotiate stage. This is good evidence of a limited degree of competition in the market from the early stages of the bidding process. The above bald figures in themselves give rather too generous a view of the degree of competition: it is clear from an examination of the detail of the relevant documentation that in a number of cases, Councils were counting bids which had very little chance of ever becoming realistic. For example, in one case, which went right through to the final selection stage, the consortium involved was clearly just a shadow company with no staff and no track record. Another way of looking at the degree of competition in the market is to consider the overall number of firms involved. The 79 bids at the final selection stage involved 22 construction firms, or groups of firms: (where firms habitually work in groups, we have counted the number of groups, rather than individual firms.) And the construction work for the 37 final contracts was handled by 20 firms (or groups). Again, this is indicative of a fairly restricted market.

Now let us look at some other aspects of PFI bidding: namely, how well do Scottish firms do in the process. On the one hand, it is clear from our data that a few Scottish firms have done well with PFI schools contracts: for example, Robertson have been involved in consortia which have secured five PFI schools projects worth almost £300 million in capital value: and as well as being involved in the construction for all these projects, they have secured the facilities management contracts for four of them. Overall, however, the bulk of the firms involved in PFI schools projects have headquarters outside Scotland. Of the 24 firms involved in the construction work on the 37 contracts, (and here we are looking at individual firms, not groups), only six are headquartered in Scotland. As regards the service element of the PFI projects, only two Scottish headquartered companies gained facilities contracts, (covering nine projects in all). The data is therefore indeed consistent with the view that PFI is not good for Scottish firms. This is in line with what researchers at Glasgow Caledonian University found, in an earlier study of the Scottish Executive PFI database.

Another claim which is often made is that PFI introduces bias into the procurement process – away from refurbishment and towards new build. It is certainly the case that, in several of the projects we have studied, initial proposals by the Council which involved refurbishment of existing schools were modified to include a greater, or even total, new build element. This happened in at least twelve cases – and in two cases this resulted in the collapse of the original project design. On the other hand, there were no cases where the change was in the other direction. Of course, we do not know that this was necessarily a bad thing: what emerged finally may have been an altogether better and more cost effective solution. But the data is certainly consistent with a bias towards new build in PFI projects.

There are also concerns that the complexity of PFI leads to a long bidding process, in the process of which costs tend to escalate. There are plenty of examples in the data of both delays and cost escalation – with, in several cases, significant cost escalation occurring after the selection of the preferred bidder. Possibly the most pronounced example is Clackmannanshire schools, where the whole process from the outline business case to financial close took four and a quarter years: costs escalated from an initial estimate of capital value of £35.8 million to £72.7 million at financial close: more than £10 million of this escalation actually took place after the preferred bidder was selected. Another aspect of the length and complexity of the bidding process is that this in itself involves high cost, for both tendering firms and Councils. It has been estimated that the cost to each bidder associated with the tendering for a large schools PPP project can be upwards of £1.5 million. Ultimately, of course, the costs of successful, and unsuccessful, bids find their way into the eventual costs of delivering services.

The findings from our comprehensive study of all PFI schools projects in Scotland confirm many of the negative suspicions that have been held about PFI, and also point to previously unidentified aspects of the schools PFI process in Scotland. In particular, the facts that Scottish schools PFI projects tend to be large in themselves relative to the norm for the UK, and also that Scotland represents such a large share of all schools PFI projects in the UK, have serious implications. Local authority expenditure on PFI projects is contractually committed, and hence effectively ring-fenced: in addition, the costs are indexed to rise in relation to inflation. This means that PFI costs are inevitably going to pre-empt an increasing proportion of local authority budgets as we move into a period of stringent public expenditure cuts. Because of the scale of schools PFI in Scotland, this squeeze is going to be much more severe in Scotland than in the rest of the UK. There is therefore unlikely to be any Westminster initiative to rescue English local authorities from their PFI squeeze, from which Scotland might have hoped to benefit through Barnett consequentials. The likelihood is that Scotland will have to face on its own the consequences of the rash implementation of PFI North of the Border.

The present Scottish government dislikes PFI, quite rightly, and has attempted to move on. But in doing so, it has made the disastrous mistake of not taking on board key lessons to be learned from past experience with PFI. One of these key lessons is that many of the adverse features of PFI stem from the sheer size of the contracts. Sheer size itself restricts competition, freezes out Scottish firms, and complicates and lengthens the bid process. And yet, in putting forward its proposals for the arrangements to replace PFI, the Scottish government did not take adequate steps to reduce the size of projects. For example, in its consultation for its proposed PFI replacement, the Scottish Futures Trust, the Scottish government included the requirement that projects involved should be ‘off the books’ in terms of the previous requirements of government accounting. However, by the time the consultation document was issued, it was already known that changes in government accounting standards were going to bring almost all PFI projects on the books anyway – so the ‘off the books’ requirement in the consultation document was pointless. But it was precisely the requirements of the old ‘off the books’ test which led to complex, bundled, and almost inevitably large PFI contracts. In other words, the Scottish Government was quite needlessly ensuring that some of the worst features of the old PFI would be carried on under the Futures Trust.

But this is not all. In those cases where the Scottish government has completely abandoned PFI or its clones, and gone back to old fashioned procurement, it has still not appreciated the lesson of size. For example, the much heralded conventional procurement of the new Southern General Hospital has been let recently as a single contract with a capital value of over £550 million: no surprise that a contract of this size did not go to a Scottish company. And still on the same theme, the present Scottish government has enthusiastically embraced the principles of the McClelland report on central procurement by public bodies. This report advocated, with the best of value for money intentions, that procurement should be bundled and centralised. Again, unfortunately, this has led to large contracts, with their resulting consequences. As a result, the Scottish Chambers of Commerce issued a statement in September, complaining that their members were being squeezed out of billions of pounds worth of public procurement contracts.

But there are other lessons too which have not been learned. Another important point about PFI and public procurement generally is that there are social and macro economic implications which go far beyond the question of value for money on individual contracts. We have already seen some of these effects above, in relation to the adverse impact of PFI on Scottish firms. But there are much broader ramifications, for example in relation to employment, training, the pool of skills in the economy, research and development, business creation and business growth. In total, the way in which the £3.3 billion of schools capital was let under old fashioned PFI by the previous Labour/LibDem administration had an adverse impact on all of these aspects. There is no evidence that the present SNP administration is taking any wider perspective in its approach to procurement – despite having plenty of opportunity to do so if it wished. For example, if it had been less interested in financial engineering and more interested in real engineering, it could have designed the Futures Trust to provide central expertise in project design, procurement and management, in order to assist public bodies to split down complex requirements into manageable individual contracts. This would have enabled Scottish firms to compete on a level playing field. And this would not be just Scottish construction companies, but companies involved in design, architecture, cleaning, catering, and other services.

Overall, the conclusion we draw is a depressing one. It is not just that we will have to suffer the consequences of old PFI in Scotland for years to come: but, as we attempt to move forward, the lessons that ought to have drawn from that experience have not been learnt.