With respect to estimates concerning the total cost of the various bailouts, etc. for the financial system, in particular whether the spending should be treated as an expenditure or an investment, Steve Waldman says:
Expenditure vs investment — thinking clearly: ...Paul Kedrosky is a reasonable fellow, and takes care to note that the numbers "are in current dollars, and all treat expenditures and investments as equivalent." Kevin Drum is even more reasonable:
This stuff has gotten completely out of hand, with "estimates" of the bailout these days ranging from $3 trillion to $7 trillion even though the vast bulk of this sum comes in the form of loan guarantees, lending facilities, and capital injections. The government will almost certainly end up spending a lot of money rescuing the financial system (I wouldn't be surprised if the final tab comes to $1 trillion over five years, maybe $2 trillion at the outside), but it's not $7 trillion or anything close to it. People really need to stop throwing around these numbers as if the bailout is comparable to World War II or something. That's not reality based, folks.
But reasonable and right are sometimes different... We have some idea what we paid for, for example, with the $851,000,000,000 for NASA. We bought space shuttles, satellite systems, a moon shot, planetary probes, a lot of research and development, some air bases and research facilities.
What are we buying when the government purchases mortgage-backed securities, or buys preferred shares of banks that can only pay if a portfolio of real-estate loans does not totally sour? We are buying "paper", right?
No. We are not buying paper. ... All of the iffy securities that are weighing down the banking system represents money already spent on real projects or consumption. When the government purchases a security, it is taking the place of the party that originally fronted money for that expenditure. Every penny of government "investment" is retroactive expenditure on housing, real-estate, consumer credit, whatever.
If a government were to borrow funds in order to build a new stadium, we'd call that an "expenditure", even if we fully expect use fees and incremental tax revenues to eventually turn a profit for the fisc. Politicians supporting the project would call it an "investment", quite justifiably. But the project would still count as government spending.
If a private party builds the same stadium, and then is reimbursed by the government in exchange for rights to future revenue, that doesn't change the economic substance of the transaction at all. But in the second case, the government would buy "paper" — it would enter into a contract trading current government funds for future revenues. That "security" doesn't make the transaction any more or less an investment than if the government had purchased the stadium itself.
So, in economic substance, the government is currently spending through a financial time machine on the exurban subdivisions and auto loans of several years past. ...
I hope that the infrastructure we build next year turns out to be a wise investment, both in financial and use-value terms. It might be, but just because we hope to recoup the cost, we won't pretend that no money was actually spent. We'll call the whole thing an expenditure, even though that will probably overstate the ultimate burden. But if a power grid counts as an expenditure on government books, so should a security derived from a mortgage or credit card loan made two years ago. You ... can't claim that securities are "investments" while a power grid, or NASA, or even World War II are mere "expenditures". ...
Figures of 7 or 8 trillion dollars recently bandied about by the Communists at Bloomberg are overstated, since they do not distinguish between expenditures and guarantees, which are contingent liabilities. The government's contingent liabilities aren't usually counted as spending until the contingency has been triggered. But the amount of money already spent or committed on "financial investments" to date is more than $3 trillion dollars, and it is perfectly right to call that government spending on the financial bail-out.
The scale of the largely unlegislated current government program to save the financial system is breathtaking and quite unprecedented. Taxpayers might be made whole, in financial terms, or might reap sufficient dividends in terms of suffering avoided to justify the program. But don't let anyone convince you that the scale of this intervention is "overstated" because it is all "investment". NASA and the Marshall Plan were investments too, and pretty good ones.
But shouldn't the example be a little different? If the private sector builds, say, a stadium and then the government buys it, then yes, that is expenditure. But suppose the government purchase comes with a clause that says it will sell the stadium back to the private sector at a date certain (or by a date certain). It's still an expenditure of the same amount in the present, but the purchase price does not represent the expected long-run burden of the transaction, and isn't that what we really care about? The government plans to sell the financial paper, not hold it forever, and what really matters is how much the paper will be worth in the future (if the stadium value falls to zero, then the current expenditure does represent the long-run burden; however, the value of the government holdings will not fall to zero or anything even close to that).
So I don't care what you call it, expenditure, investment, a repo, temporary custody of a volatile asset, whatever, what I care about is how much the bailout will cost once the government has disposed of all of the assets it has purchased. That's not something we can know with certainty, but unless the value of the securities the government is holding falls much, much further than anyone expects, the amount of the current expenditure greatly overstates the long-run burden.