Brad DeLong has a piece for on “debt” at The Milken Review
https://www.milkenreview.org/articles/federal-debt?IssueID=53
And Noah takes another bite from the same apple at:
So “debt” must be a thing.
Both DeLong and Smith, of course, have totally correct analyses. [I’m going to discover an error in Brad Delong’s or Noah Smith’s economic analysis?]
Even though:
The US government provides a unique means of safekeeping the assets of corrupt public officials and nervous nellies the world over.
And
Debt/GDP ratios are not a definitive constraint.
Nevertheless
The USG cannot increase the deb/income ratio indefinitely with no downside.
Still, both are asking what level or trajectory of debt avoids a debtpocalypse. This is, however, not the right question. Debt is neither an outcome (an argument in the utility function anyone but Ken Rogoff 😊) or a policy instrument that affects any outcome variable. Debt is just the arithmetic sum of deficits, which in turn is the algebraic sum of the differences between expenditures and revenues going back to the Big Bang.
For sure there is _some_ relation of Debt to outcomes. Debt affects the interest rate at which the government can borrow and thereby indirectly affects the discount rate of the NPV of any given expenditure (and even more indirectly the dead-weight loss of any tax). Fewer expenditures will pass that test and the value of more taxes will exceed their dead-weight loss when debt is high. This implies (or ought to imply) that at higher Debt to Income ratios non-debt service expenditures will be lower and taxes higher.
But if all expenditures DO pass the NPV test then the debt to repayment capacity will fall so the borrowing rates will fall as well. Therefore, an equilibrium rate of increase in the debt and an equilibrium debt-to-income ratio exist. Note, however the key assumptions that all expenditures pass an NPV test and taxes are large enough to cover the interest on the debt. Faster increases lead to Delong/Smith debtpocalypse.
However, the debtpocalpse-avoidance equilibrium rate of debt to income is not necessarily the rate at which real income growth is maximized. At higher debt to income ratios more expenditure should be financed by taxes than by borrowing. Higher taxes to establish a lower equilibrium D/Y ratio might be better for growth.
If the NPV of expenditures <0 even if they are financed with the optimum combination of taxes and borrowing so as to avoid debtpocalypse, still have costs in slower growth. It is in this sense today’s deficits are a burden on future generations. Conclusion: it is better to focus directly on trying to optimize expenditures, ensuring that the NPV > 0 and since that is unlikely, finance any expenditure with NPV < zero with revenue from low deadweight loss taxes, than to worry about controlling debt to avoid a debtpocalypse.
Sorry Ken.
[Standard bleg: Although my style is know-it-all-ism, I do sometime entertain the thought that, here and there, I might be mistaken on some minor detail. I would welcome comments on these views.]
Good post. I do disagree though that debt is not a policy tool. Debt is an inter-generational savings tool. In a perfect world (from an efficiency stand points), we would have a direct transfers (in the vein of Social Security). Since we don't, we have to use the second best tool, which is a savings tool called government debt. The tool enables us to do the NPV positive investments.