Money isn't static. We only come out ahead if that money in GM is better than any other investment that could (or would) be made. This is unlikely. Put another way, if/when we hit a break-even point, it will be as if we had taken $82 billion dollars and hit it under a mattress for three years, rather than putting it to some other worthwhile economic use.
There's really no way around the fact that all of these bailouts and stimulus programs hurt. The fortunate part is that they are all as cheap as they could be because they've managed to serve their purpose of keeping vital industries afloat and US debt interest is running below inflation right now (meaning that there's actually a slight profit to taking out debt).
There's only a profit to taking out debt if the investment has a multiplier that is equal to the savings rate of inflated debt (using 5x5 inflation expectations with a savings of 55bps a year that number is .9467). Most government projects unfortunately have a multiplier less than one, meaning that even with negative real rates the investment costs more than the fundamental value. This is known as the overhead effect in economics. Projects must generate positive NPV including all the costs of administration and fees associated with the project.
Simple example. If the Federal government decides to send out $1 to every citizen, it costs more than $1 to accomplish this. There is the cost of postage, the cost of making the decision, etc. The same takes place in a business. That is how you get a multiplier less than one. i.e. 1/real cost so will use .7 as the example. The project would cost 1.42 to get every dollar into the economy, meaning the multiplier is .7.
Even though we are paying economic rent (interest) that is less than the inflation rate on borrowed principal, the difference must equal the real rate differential as well as the multiplier difference in the project.
In the above example, assuming inflation runs at 55bps more than the 10 year nominal rate (using 5x5 inflation breakevens) the above project still costs ~3.0% per annum in real economic cost the savings.
The major problem is that many government projects don't generate positive multipliers. Real interest rate differentials can't fix this.
So no, we aren't doing ourselves a favor in borrowing as it is cheaper than 5x5 inflation breakevens.
(And again, don't take this as me making a position either way on the large fiscal deficits, just more or less trying to state facts as I like to watch you guys argue.)