Spending, Money, and Time, Part 3
… wherein I find myself in the totally bizarre world of siding with Paul Krugman against Nick Rowe. However, I don’t think there is a real disagreement, just a misunderstanding. I will stand by my claim that consumption cannot be moved, in aggregate, without increasing/decreasing natural resources and real investment. That does not mean, however, that there cannot be intergenerational transfers. Government debt is a strange creature which breeds much confusion. Let’s go through what happens when the government borrows money to spend:
GDP = $1,000 forever, Prices are perfectly flexible, no capital/natural resource depletion, economy is permanently producing at capacity, 0% interest rates, and all pulleys are massless.
Time Period 1:
Government borrows $100 from people and spends it. People at Time 1 spend $100 less, government spends $100 more. Crowding out can be achieved through higher interest rates, price adjustment, or whatever.
Time period 2:
Debt comes due. Government increases taxes by $100 and uses proceeds to buy back (and destroy) the bond. Negative income effects from taxation are balanced by positive income effects to bondholders. Private spending is back to $1,000, minus the deadweight losses from taxation.
Thus the debt causes the reduction in private spending when it is issued, not when it comes due, under these very unrealistic assumptions. So, one could think of bonds as a way to defer the deadweight losses of taxation. The largest cost of the government spending, however, is still in time period 1, where the government spending crowds out private spending. People in time period 1 do not get more, they simply get more government and less everything else.
What if the money in time period 1 goes to the old? Well then, the young consume $100 less and the old $100 more in time period 1. If you want to call that an intertemporal transfer, fine, but I don’t think that’s what Krugman means. I think he means that government spending now crowds out private spending now, except he can’t directly say that because his left leaning readership would flip out, so he words it in a strange way so they won’t notice. When people like David Brooks say that debt burdens hurt our grandchildren, I think he means that somehow our debt now will cause our GDP to be significantly lower in the future, not that if you have an upwardly spiraling welfare state, then children tomorrow will have to pay more for elderly tomorrow. When tomorrow comes, if the welfare state is unpopular, the government then can abolish it. If deadweight losses from taxation to repay debt become too high, the government can default.
The Taxation/Debt Margin
In Rowe’s model, government debt is a burden to our children because the government is primarily a welfare program for the old. It’s not the debt that burdens our children, it’s the transfers. If government spending were done on all cohorts equally, there would be no transfer effect.
The same effect can be achieved with 0 debt, just taxation.
Period (n): Young are taxed $100 to give to the elderly.
If you want to have it eventually explode, add:
Period (n): Young are taxed $100*(1.1)^n to give to the elderly.
Note that while each generation gets a transfer payment, at no point in times does consumption in one time period ever exceed past production. While increased intergenerational transfers can occur, they do so by burdening the current young generation with crowding out and higher deadweight losses. In conclusion, it’s not the debt that causes the burden to our children, its the welfare programs.
Here are links to part 1 and part 2.
Noapinion on Ricardian Equivalence and Say’s Law.
Barro “Are Government Bonds Net Wealth?”
Arnold Kling recognizes debt is both an asset and a liability.