I found the advice from the inspector-general of taxation, that the Australian Taxation Office (ATO) should avoid long, fixed-cost projects where one vendor has the reins, to be a bit short-sighted.
The argument was that with a fixed-cost and single vendor, there was a risk that scope would change and as timelines began to slip, the vendor would drop the quality of the end-result so that it could finish at an acceptable cost.
Therefore, projects should be passed out in modules, smaller bite-sized chunks that would have lower risks of expanding in scope.
Talking to IBRS analyst Jorn Bettin about this concept, he believed that its success would depend on how the responsibilities were divided amongst vendors. There can be a lot of finger pointing going on when multiple vendors are working on projects that dovetail together. That's why companies often like "one throat to choke".
Bettin thinks that to have multiple vendors working on small modules for an overarching program, their work should be divided by competencies.
But that causes its own issues, according to Bettin, because it's difficult for the government to truly understand what big vendors' competencies are. Meanwhile, small vendors, which may have well-defined competencies, are passed over because of their risk profile. If something goes wrong, will they be able to make it right or pay to fill the gap?
My biggest issue with having multiple vendors is that even if work is split into competencies, it requires the government to be a real leader, guiding the vendors into their individual berths like a harbour master. Past experience has shown that government project management and IT governance leaves much to be desired, and I could see such projects ending in absolute shambles.
So I'm afraid, in this case, it's probably damned if you do and damned if you don't.