"...cost taxpayers £10.7bn... ...was anticipated to cost £4.1bn...."
Aspire is a services and procurement framework, not a singular contract. When it was originally spec'd the expected spend was approx £4bn. Actual services procured through the contract are the £10.7bn figure. It's not that it's 2.5x over budget, more like HMRC have bought 2.5x more than they expected over the time frame.
When you bear in mind that time period (c. 10 years) has involved the delivery of online self assessment, RTI, tax credits and a whole raft of behind-the-scenes analytics programmes that shouldn't really surprise anyone.
Which is why this will be closely watched (and why PAC are concerned). Last time around HMRC underestimated their own services by more than a factor of two. This time round they've got the pleasure of having to do it with no in-house project management or procurement capability. One little known fact of Aspire is that until very recently all PM and PMO activity was handled by Capgemini (i.e. Capgemini managed Capgemini contracts, heh). Last count had the number of active agreements within the framework at something like 17,000.
Between novation and this tiny little TUPE they've taken over all of 10.
2017 is the end date for Aspire. Yeah, right.