When I do the math....it doesn't come out..what am I missing.?

Hmmm... post your cost breakdowns and I'll have a look at them. Seriously, though, what you're probably missing is experience in managing development programs of this type. There are always unexpected costs when dealing with cutting edge technology. Also, the reliability requirements for space probes adds a lot of cost for testing, certification, redundancy, etc.

That said, the Webb Telescope has lots of people wondering how it got so far over-budget. That program has been plagued with delays and do-overs.

The cost generally cited for the Curiosity program is $2.5 billion; where did you see $4B? Your under $1B estimate is definitely lowball - the launch vehicle alone was about half a billion.

:-)

There used to be a joke that, when the government (or pseudo government organisations) are involved, the cost of a hammer that you can buy for $2 is really $2,000.

How? Well, worker buys the hammer for $2. He bills the company for the tool. The company charges its clients cost plus 10%, so the cost to the client for the hammer is $2.20. The client charges its client cost plus 10%, so those clients pay $2.42 for the hammer.

When you have 10 layers, that means $2 *(1.1)^10 = $5.19 or $5.20. Put in enough layers, and it's like compound interest.

Note that NOBODY really overcharges for anything -- that hammer is just one little bit of the whole bill, but it can finish up costing way more than it would down at the person level.

It's the cost-plus mentality of most bureaucracies that causes the problem.

Oh, and as for NASA people being underpaid, in some ways they are; most of NASA's budget goes out n two ways -- administration and to contractors who hire contractors, who hire contractors... With the compounding, it doesn't matter that no one is overpaid, And most times, the cost is cost plus 50%...

Note: some consultancies work at cost + 140%....