Last lessons from the Harmon affair

Four years ago, both of us were having World Cup results fed through to us in court in the middle of the High Court trial Harmon vs House of Commons. The case concerned a claim by Harmon that it should have been awarded the fenestration package for the new parliamentary building, now called Portcullis House. The cost to the taxpayer has been £10m.

The National Audit Office finally produced its investigation into the project. An earlier report concluded that serious mistakes were made by the civil servant in charge, even though he was acting in good faith. It is interesting that the first inquiry should clear him. His Honour Judge Humphrey Lloyd described his conduct as "not the actions of an honest or reasonable man". It would be interesting to know how the same events came to be described so differently.

The NAO report raises questions of wider importance regarding the purpose and relevance of tendering in complex projects. The House of Commons' tender procedure was supposed to accord with European Union tender regulations so that the tenders were made on a common specification, and judged against stated, objective criteria. In fact, the successful tender was made on the basis of a design that none of the other tenderers were asked to tender on, and the other tenderers were not given a chance to put forward their own variant bids.

Clearly, if the purpose of open tendering is to ensure that tenderers have the opportunity to have their offers compared fairly, then tenders must be submitted on a comparable basis. One of the recurring problems in procurement law is to decide when the contract or specification has changed so much from the basis on which the tenders were sought that it has become a new contract or specification – in which case the client ought really to be starting a new tender procedure. This problem is particularly acute because, under EU law, a contracting authority is not, in theory, permitted to enter into price-sensitive discussions after tenders have been submitted.

The NAO report puts the best face on the mistakes made. It explains that the starting point for these discussions was when the tenders for this package came in more than £22m over budget. A group of consultants was established to reduce the cost of the package while "preserving its fundamental design principles".

The House of Commons scored an own goal when it decided in the course of this process to ask for three tenderers to submit tenders on the basis of one design solution, and the other (eventually successful tenderer) to submit a bid on the basis of another design, which it had already decided was the preferred design. It is hard to see, in those circumstances, what the tender procedure was for other than to provide the appearance of compliance.

The NAO does not come out and express any clear view as to whether the outcome of these discussions with the successful tenderer achieved value for money (or at least might have done if the House of Commons had not had to pay the extra £10m for the court proceedings). The sense from the report is that it did – but if that is the case, it raises serious questions about the relevance of open tendering to achieving value for money.

Sir John Egan would say that it is the government's attachment to tendering that makes it a "bad client" (see Building, 19 April, page 13) – and he is right that in so far as public (and utility) clients are subject to EU procurement legislation, they will be wedded to tendering. That is required by law.

The real difficulty is that although discussions between clients and contractors may have a number of advantages in achieving Egan principles, if there are no other truly comparable tenders to refer to, there is a substantial risk that value for money is established only by reference to the sort of "pseudo-scientific mumbo-jumbo" that the NAO has recently said is often used as the basis for justifying PFI projects. If public clients cannot be relied on to achieve value for money by entering into contracts following negotiations with a single tenderer, it will surely be inevitable that the discipline of the tender procedure will be imposed on them.