Paul Krugman says that now is no time to worry about deficits:
Just to be clear, I’m not arguing that trying to reduce the budget deficit is always bad for private investment. You can make a reasonable case that Bill Clinton’s fiscal restraint in the 1990s helped fuel the great U.S. investment boom of that decade, which in turn helped cause a resurgence in productivity growth.
What made fiscal austerity such a bad idea both in Roosevelt’s America and in 1990s Japan were special circumstances: in both cases the government pulled back in the face of a liquidity trap, a situation in which the monetary authority had cut interest rates as far as it could, yet the economy was still operating far below capacity.
And we’re in the same kind of trap today — which is why deficit worries are misplaced.
One more thing: Fiscal expansion will be even better for America’s future if a large part of the expansion takes the form of public investment — of building roads, repairing bridges and developing new technologies, all of which make the nation richer in the long run.
While on this subject, it’s worth saying that infrastructure spending really ought to be treated differently for deficit purposes than other kinds of spending. Most government spending — on defense, on Social Security, on Medicare and Medicaid, on food stamps — is basically analogous to household consumption. Here you want the accounts to balance out over the long run. It’s fine to run a small operating deficit during a normal recession to avoid disruption, and a big operating deficit when you fear a liquidity trap to avoid depression, but you want to keep the long-run debt load in check so you want balanced budgets or even small surpluses at other times. But infrastructure spending isn’t like this and shouldn’t be treated that way. It’s more analogous to a business’ capital spending. If a business builds a new factory, it doesn’t count all that spending as “spending” and take a huge one-year loss followed by years of extraordinary profits. Instead, the cost of building the factory is spread out for accounting purposes across its useful life on a depreciation schedule. That provides a more accurate picture of the year-to-year financial health of the business.
Infrastructure ought to be accounted for in a similar way. If you’re considering building a bridge that will take 2 years and $X to build, but that will last for decades, there’s no particular reason to think you should raise enough tax revenue to cover the cost during that two year period. Borrowing money in order to make investments that have long-term payoff is what borrowing is there for and the government can borrow on unusually good terms. Of course borrowing money to finance stupid infrastructure projects is a bad idea, but borrowing money to finance beneficial ones is fine even under circumstances when borrowing for other purposes wouldn’t be.