Until you realize the negative feedback loop of hundreds of thousands of GM employees hitting unemployment lines, and the secondary and tertiary companies that supply and contract for GM also shutting down. 200,000 employees * 1,000/mo unemployment * 100 weeks = $20 billion dollars. Obviously a terrible case, but not even a worst case as I examine only GM employees and not the subsequent losses in direct suppliers/contractors and community losses relying on those employees patronage.
Sure, the investment may not have great returns.
But are you calculating the billions in reduced welfare, or the billions in increased GDP, or the effect of stopping a negative feedback loop which spirals more deeply?
I mean, a GM level destruction isn't just GM. Isn't just suppliers. It's grocery stores. It's car mechanics. It's big box stores. It's fast food. It's everything in those communities that former employees patronized.
When you examine the decision more deeply than "net dollar returned" I think you'll develop a greater appreciation for the economic benefits of entire communities not being ruined.