Let's hypothesize that all the engineers working in SV get an AIDS-like illness caused by spending too much time on news.yc.
If there was a health care solution that would make sure that all the employees of Google, Intel, HP, etc. continued to function as usual by caring for those engineers, then even an extremely high up-front cost for that care would result in a net-win for the budget/economy/country.
Of course, that's silly as it won't happen to that extreme. But, I don't think the downward-spiral of more-cuts-here-and-there that you describe is the only possible outcome. A positive virtuous-cycle of increased wealth production is also possible.
The only ways that increased healthcare spending can increase productivity are through effective preventive measures and corrective treatments that return the patient to sufficient health to be productive, but that's not where our super-high healthcare costs come from.
They come from the consequences of delayed care (eg. using only emergency rooms when you're uninsured, necessitating more risky treatments for more advanced problems, which means higher malpractice premiums) and from end-of-life care.
Nice reply, btw :)
If you instead look at all government spending as economic stimulus, and all taxation as a factor in inflation management, you get quite a different picture.
It can be carried further: money they get refunded comes out of other areas' funding, as the difference gets made up somehow. Their property taxes go up, the cost of food goes up because others are carrying a heavier burden, jobs across the board pay less / hire fewer, everything always costs something, even refunds. Less in some situations, but not zero. And we're still not reaching $50-$100.