Ain’t It Fun, Living in an r < n+g World?: Tuesday Focus for September 9, 2014

Consider r, the required real rate of interest on government debt, and the sum of n, the labor-force growth rate, and g, the labor productivity growth rate.

When r > n+g, the case for budget surpluses and lowering the debt is strong. We believe that private investments are even more productive to society than the market rate of return on private capital formation tells us. We believe that there are private knowledge and other spillovers from private capital formation. We believe that the quasi-rents from past investments are shared among all those with market power who participate in production. Thus when the government borrows and spends, it crowds out private investments that are very profitable to society as a whole. This point is strengthened by the danger of a runaway government financial crisis should a default premium begin to incorporate itself into government borrowing rates. This point is limited only by intergenerational equity considerations: that if we think the future will be richer than the present, committing funds now to private capital formation rather than to private and government consumption worsens intergenerational inequality.

The problem is that we really do seem to be in a different world than the r > n+g world–in fact, we seem to have been in a different world since 1933. And the question of what to do when r < n+g is much harder. There are then, I think, three possibilities:

One possibility is that r < n+g is telling us that private returns from investment, when properly adjusted for risk, are truly less than the growth rate of the economy–that the economy truly is dynamically inefficient. In this case, the reduction in private capital formation from higher government debt is not a bug but a feature: the economy does have too much private capital, and less would improve welfare.

A second possibility is that private capital is still productive and valuable, but that rent-sharing by stakeholders and the inability of investors to capture knowledge externalities pushes the risk-adjusted private return on capital and thus the government borrowing rate below the total economic growth rate. In this case there is a case for adding a cost-of-crowding-out term to the benefit-cost analysis of government debt finance. But it is also important to recognize that there is no danger of a runaway government financial crisis should a default premium begin to incorporate itself into government borrowing rates.

The third–and I think by far the most likely–possibility is that the fact that r < g for the government is a byproduct of an extremely large outsized risk premium because of private financial markets’ failure to mobilize the risk-bearing capacity of the public and failure to establish trust and overcome moral hazard in the credit channel. Thus more government debt provides the private sector with something that it is willing to pay through the nose for: a low-risk way to transfer purchasing power into the future.

Under such circumstances a higher government debt does not reduce but raises the amount of programmatic spending that can be sustained by any given tax share of GDP: an extra proportion (n+g-r)(D/Y) of national income is available each year for government programmatic spending or tax cuts. Plus there is the direct utility gain to debtholders for providing them with something that they value greatly. In this case, there is also (a) no reason to worry about exploding government debt and a consequent government debt crisis, (b) a relatively strong case for using the government’s borrowing power to route around the private market’s inability to unclog the credit channel and mobilize risk bearing capacity, hence (c) little reason to worry that government debt is crowding out important private capital formation–for, after all, the credit channel is badly clogged.

I certainly think that the evidence is sufficiently ambiguous that we should not make big and irreversible policy moves on the assumption that we do not live in an r

And the large long-run fiscal gap stems (a) from health-care costs exploding much faster than in other countries, and (b) from assumptions that current law provisions that restrict spending growth and raise taxes will be overturned by future congresses. It seems incoherent to say that we must do more short run damage to the economy by passing more current laws to control future deficits that can then also be overturned by future congresses.

As I look at it, there are three things not in the CBO’s Alternative Fiscal Scenario that lead me to think that we did not have a large long-run fiscal gap: (1) That someday, somehow, the rise in health-care costs would moderate and we would move toward the OECD average on health spending. (2) The Cadillac Tax and the resulting shift of what are now untaxed employer health benefits into the tax base. (3) The carbon tax–which will come sometime in the nest generation. It seemed to me that an AFS or a generational accounting calculation that did not include these was more a tendentious political intervention in the current debate then an honest technocratic forecast,

And, of course, after Reagan 1981 and Bush 2001 I have to believe that any laws passed that restrict long-run spending growth when Democrats have influence will be undone by laws that cut taxes in the long run when Republicans take control of the government.

The arguments for austerity or even for stabilizing the debt-to-GDP ratio thus seem to me to be very shaky. They are claims that:

1. Soon, very soon, r is going to rise to be greater than n+g, and much greater than n+g–but when it does we will be unable to change our fiscal policy from one appropriate for r < n+g to one appropriate for r > n+g by a substantial margin.
2. Even if r < n+g, there is still a social loss from the crowding-out of private capital formation that takes place when government debt is increased.
3. With taxes fixed as a share of GDP by political pressures, and with deficits limited by deficit phobia, a higher debt means lower programmatic spending.

(3) seems to me to be simply confused. (3) says that the present value of spending is fixed, so all that spending increases and cuts do is move spending around from decade to decade. In that case, you would want spending to be high when r is low because the amount of future spending you have to forego to stay under the tax + interest cap is low; and you would want spending to be low when r is high because they you are foregoing a lot of future spending for each dollar of spending financed by debt today. It’s not an argument for not running a large deficit now, when r is low. (2) is possible, but needs to be developed more. (1) is an argument that we should not even set out what the best policy is now because our current political system is so broken that even pointing out that r < n+g will lead to destructive political economic consequences.