What is often portrayed as an "overspend" (so the Daily Wail fraternity can have their daily dose of hurrumph) is in most cases either one or more of:
1. The increased cost associated with changes in the requirement which are themselves simply the result of the facts that (a) we do not live in a static world, and (b) even the most dedicated servicemen and women do not have calibrated crystal balls. These costs are identified, challenged and justified in a pretty rigorous manner as a matter of course.
2. The detail that the original cost has been taken by the press (and cynical people with axes to grind) to be "firm fixed prices" when they are usually either ROM prices or the raw cost before Risk. As a project progresses some of the risks occur and so the *spend* rises above the baseline *cost*, but only in so far as the original risk cost was not included in the press report. This is all understood by the people who matter - the contractors, the MOD (uniformed and civilian) and the Treasury, but is frequently wilfully misrepresented to the public by people with axes to grind.
3. The consequence of naive contracting - there have been cases where to reduce the "headline price" the MOD has elected to take ownership of all risks and only contract for the minimum baseline cost. This approach is actually allowed by the guidelines, but only where the contract has an initial "Risk Reduction" Phase before the full funding is committed. There have been times where this approach was taken but the MOD refused to have a Risk Reduction phase - MRA4 was a classical example. In this case the "cost overrun" would actually be more accurately described as "the true cost of what they were trying to achieve becoming apparent after contract award". If the contract has not been sold the risk content then the costs of those risks occuring vest firmly with the MOD - that's kinda the way it works.
4. Someone somewhere screws up. It happens, but in my experience it's comparatively rare. Even in these cases it's usually someone trying to do something for what they perceive to be the best of reasons, but not appreciating the magniotude of the task or its ramifications. The glass-cockpit Chinook and one service-engineered weapon system upgrade on the Jag would be examples of this. BAES screwed up to some extent on the MRA4 and Astute programmes, which is why they took those £800m losses on those projects.
But I would contend, based on experience from the inside, that in >>95% of the times where Something Goes Wrong in the UK it is the result of people trying to do the right thing but failing rather than neglect, incompetence, fraud or shady dealing. And even in those cases it is very often the case that the "Error" is only "obvious" with 20/20 hindsight - a detail which is invariably glossed over by the hecklers who criticise the work of others but have never actually delivered anything themselves.
€6.66 supplied, YMMV,
PDR