As a fan of reforming public sector accounting practices, I was pleased to see the following from Matt Yglesias:
The federal government, somewhat unfortunately, doesn’t engage in proper balance sheet accounting. So if you borrow $10 million to do road repairs, that shows up as an increase in indebtedness. The right way to think about it, however, would be that the road is a depreciating asset that has a value. Every year it goes unrepaired, its value declines. So borrowing $10 million to fix the road might improve the government’s net financial position, depending on the nature of the project. Right now we’re at a time when it’s never been cheaper to finance federal borrowing. Consequently, borrowing money and spending it on sound projects of long-duration is the best way we have to “save” for the future even though it technically adds debt. Now of course spending money on something dumb doesn’t help. But deferring repairs and useful investments for the sake of borrowing less is going to leave us poorer in the future rather than richer.
In our post-crisis policy conversation, we have tended to talk about stimulus vs. spending discipline. The problem, however, is that “fiscal stimulus” aggregates many different kinds of spending. If our primary concern is to stabilize consumption levels, cash transfers look attractive relative to infrastructure spending, which tends to take a long time. But there is another way of looking at these slow-motion expenditures on big-ticket, durable things: in a depressed economy, buying certain things that we are likely to need in the future is cheaper than it is in a strong economy, as there is less competition for various resources. So it is not crazy to buy these things now, e.g., military equipment, infrastructure, etc. Proper balance sheet accounting would tend to strengthen this case.