Something rotten with the state of procurement

There must be something fundamentally wrong with the procurement of construction work. A new report from The Royal Incorporation of Architects in Scotland (RIAS) summarizes a twelve-month review of public procurement in Scotland. They cite work carried out by others to highlight the huge expense of construction procurement, being the sector with the highest procurement costs across Europe. They also cite evidence that procurement is slower in the UK than anywhere else in Europe apart from Greece.

There seems to be something fundamentally wrong with the connection between theory and practice when it comes to transaction costs. Empirically speaking, if the costs of contracting-out are unreasonably high, then organizations would bring operations in house and carry them out within the organization, rather than going to the market. In the long run, assuming that the production costs would be the same in either case, the most efficient form of transaction would be the most competitive, and therefore would predominate. What, then, should we make of the calls for reducing transaction costs in construction? If they are too high, then why is construction not routinely an in-house activity?