October 13, 2012
After having provoked a debate that subsequently involved Nick Rowe, Brad DeLong and Paul Krugman, I will assert blog owners’ privilege and throw out some summary thoughts.
First, we all seem to agree that in a situation where the economy is clearly operating well below its potential, governments can run deficits to boost employment and output. I believe we all agree that in principle the government can also use these deficits to increase future output through productive investment in either physical or human capital. This would make future generations better off on net as a result of deficits today, since the economy will be larger than it would be without the deficits.
I would also add, without necessarily implicating anyone else, that simply by increasing output and employment the government is likely to make society better off in the future for two reasons. First, by keeping people employed we will keep them attached to the labor force and reduce the number of hard core unemployed who would be difficult to re-employ in subsequent years, possibly leaving us with a higher rate of unemployment (and lower output) long into the future.
The other reason that short-term increases in employment can have long-term effects is that by keeping families intact, children are likely to have better upbringings and do better in school. This means that the next generation will on average will have happier more productive lives because we used deficits to keep their parents employed today.
Okay, but even if deficits today don’t reduce output tomorrow, Nick Rowe argues that by increasing the wealth of some members of the current generation (those who hold the bonds used to finance the deficit), we can still be reducing the wealth of members of future generations. The argument here is that if we get back to full employment at some point in the future (this is a full employment argument), the people who hold the bonds will be able to pull resources away from young people who are entering the labor force and were not involved in the decision to run deficits today.
There is some validity to this point, but it is extremely limited. First, it is important to remember that most of the holders of current debt will be people who have children and/or will bequest some of their wealth to their children or charitable organizations. While children may not benefit one to one from the consumption of their parents, surely they benefit in part. Insofar as people with or without children save some of the wealth from the bonds and subsequently will it to children or charitable organizations, today’s deficit will not crowd out any consumption of future generations.
Finally, much of the tax burden of paying the debt service on the debt issued today will be borne by members of the current generation. In that sense it is an issue of intra-generational distribution, not intergenerational distribution.
In short, we can talk about some burden on future generations as a result of debt issued today, if it goes to unproductive uses, but the size of this burden is clearly a fraction of the debt and likely to be a very small fraction in my book. Furthermore, debt is far from the only mechanism through which the government imposes inter-generational burdens of this type.
Let’s take monetary policy, the preferred mechanism of Rowe and others for sustaining full employment. Suppose the Fed adopted a nominal GDP target. To hit this target suppose it bought up some massive amount of long-term bonds, both long-term Treasury bonds and mortgage-backed securities (MBS). This could be expected to drive interest rates even lower than they are now.
Suppose that the interest rate on 30-year Treasuries falls to 2.5 percent. Let’s assume that this sparks enough demand (along with expectations of higher inflation) to get us back to full employment. If this plan goes according to plan, then in 2-3 years’ time the economy will be more or less back to normal with moderate inflation and unemployment back to a 4.5-5.0 percent range. The Fed will then be looking to raise interest rates to keep inflation from rising too high and thereby exceeding the nominal GDP target.
This would mean that the price of the long-term Treasuries and MBS will have plummeted. When the Fed sells off these assets to reduce the money supply and thereby raise interest rates, it will take large losses. In this story, it would likely need additional tax revenue from the Treasury to cover these losses. That would imply a tax burden on future generations in the same way as the formal debt that so concerns Rowe.
The winners in this story are the homeowners and others who were able to borrow long-term at very low interest rates. They will be able to consume more in the future as a result of these long-term low interest loans, with their consumption coming at the expense of the consumption of future generations. Shouldn’t we feel every bit as bad about this story as the story of the deficits that upsets Rowe? (We get a similar story if low interest rates push up house prices, thereby allowing current homeowners to cash out with big gains.)
There are a few other points worth making in this story. First, there is no magic to zero. Any government spending that does not have the character of investment can be seen as coming at the expense of future generations in the sense that concerns Rowe. Even if we had a balanced budget we could still say that if we reduced government consumption or increased taxes, we could have a lower tax burden in the future. It’s not clear what magic zero holds in this story.
The deficit is also a tricky target. We often get lower deficits through asset sales of different types. If we balance the budget by selling off highways or parklands to private businesses that then charge for their use, it’s hard to see how we have helped future generations. The same applies to selling off the airwaves or leasing mineral rights on government properties.
The most important selling off assets along this line does not involve formal sales but rather takes the form of patents and copyrights. The government is effectively paying people to innovate or do creative work by giving them legal monopolies in certain markets. The discounted value of these monopolies likely dwarfs the value of the outstanding debt. In pharmaceuticals alone, the annual cost of patent protection is close to $250 billion a year, 50 percent more than net interest on the debt. These payments will involve much larger generational transfers of the type that concerns Rowe than the debt.
In short, there are many good reasons why we should be doing everything we can to push the economy to full employment. I am inclined to believe that fiscal policy is the most effective route, but I am happy to push monetary policy as far as we can. We seem to agree that there is no reason to believe that deficits in the current context will cause the economy to grow more slowly in the future and many reasons to believe that it could increase growth.
There is a limited sense in which the wealth created today can allow members of the current generation to consume more at the expense of younger people who will be reaching adulthood during the lifetime of those who buy bonds today. However the extent of this intergenerational transfer is likely to be small relative to the size of the stimulus and is a feature shared with many other policies where the issue is not even raised.
In short, sorry kids – you haven’t given me a reason to oppose stimulus.
Comments