As Yglesias points out, 5 year, 7 year, and 10 year US Treasuries have such low rates that their real interest rates (that is, the interest rate after inflation is taken into account) is negative.* Which means that the US government is in a better position than someone who can borrow money for free; investors are actually paying the US government to borrow their money.
This means that, at a minimum, the US can borrow a bunch of money and just let it sit there, and that would be better than not borrowing any money. But more reasonably, it means that for any expenditure that we know must take place within the next 10 years, it makes sense to pay for it now rather than later (if possible).
The most obvious category of spending that presents itself is infrastructure. Over the next decade, we know we're going to have to build new schools and hospitals and upgrade old ones; expand airport capacity and update our air traffic control systems; repave roads and replace old pipe and sewer systems; build new highways and bridges; dig new tunnels; retrofit a lot of old buildings; expand existing rail and other mass transit; improve energy efficiency to eliminate dead-weight loss, for example by painting roofs of buildings white; etc. So there's a lot of work to be done. But there are three reasons why infrastructure spending now is a particularly good idea:
- We know with certainty that the work must be done over the next decade anyway
- Much of the work is done by state and city governments, and so federal dollars would be providing a timely boost to ailing state and city budgets
- These projects will efficiently stimulate overall demand because they will mobilize the huge amount of idle labor in the construction sector, which was hit hardest by the bursting of the real estate bubble; so there is no danger of inefficient "crowding out"**
* Say I borrow money from you at 5%. I repay you in a year--but money by then has become 5% less valuable, due to inflation. So in real (as opposed to nominal) terms, I paid 0% interest. So inflation works as a discount on borrowing money.
** In a recession, by definition, there are a bunch of idle resources in the country--perfectly good employees and machines, but no demand that puts them to work. So the way to end a recession is to increase demand so that idle resources are mobilized into productive action. One way to do that is to encourage the private sector (companies and consumers) to spend money, for example by making it cheaper to borrow money (lowering interest rates), increasing the money supply (increasing inflation, which as discussed in * has the effect of lowering real interest rates), or just giving people extra cash and hoping they spend it (tax breaks). But if that doesn't work, the federal government can just employ the idle resources directly by spending a bunch of money, say on new bridges or F-18 jets or something (stimulus). However, what happens when the government spends a bunch of money in non-recessionary times, when there are not a bunch of idle resources? Well, in the case of construction say, the government would be bidding with private entities (companies and consumers) for the scarce supply of construction labor. The government would end up bidding up the price, thus pricing out of the market a bunch of private companies and consumers--in other words, "crowding them out". Crowding out might be bad if you think the government doesn't do things as efficiently as private entities do; but this concern is moot if no crowding out is happening, which is precisely the case when the government hires labor that would have been idle anyway.
No comments:
Post a Comment