The easiest way to save cost is to build something the right the first time and not change the design and/or requirements every year or two. JSF studies started in 1993, was awarded to Lockheed Martin in 2001 after competition testing and 16 years later it’s still in test.
Have you ever worked on a DoD development contract?
Costs always increase, often dramatically, and the reason is requirements creep, which is usually a government issue, not a contractor issue. This is normal. It is a development contract. New things are discovered as work is done, and are incorporated into the design.
Another problem is that the length of the contract is such that procurement personnel in both the government and the contract sides turn over and are replaced during the process, which leads to a loss of institutional memory and extreme rigidity on some requirements. Just as a hypothetical, suppose there is a range specification for the aircraft. Say 1000 miles. We get six years into a contract and suddenly the engine is found to be slightly less efficient than was believed when the program started. This means the range falls to 990 miles. Well, it turns out that you can't just add a couple of gallons to the size of the fuel tank on an aircraft, and keeping the original range requirement is going to mean a major redesign and an enormous cost overrun. Sometimes the government will say that there is no practical difference between 990 miles and 1000 miles, and sometimes they will insist on meeting the original spec.
Not quite so easy to determine fault there is it?