Deficits and Interest Rates

Asymmetrical Information's newest contributor, Winterspeak, argues that at present the US budget deficits don't matter all that much, in part because interest rates are too low anyway:

This Badness would not be outweighed by the traditional Goodness from small budget deficits -- namely lower interest rates. The US short term rate is currently being actively increased by the Fed because I assume they think it is too low now and possibly feeding into a dangerous real estate bubble. The long term rate is arguably too low as well as Asian central banks take money from their people and give it to the US.

That's not his whole argument, and certainly not the best part of it, but it's the only part on which I plan to comment, so read the whole thing if you want to know more.

It's always seemed to me that saying that deficits do their damage by raising interest rates is a very roundabout way of explaining the problem---one which tends to confuse the issue and lead to questionable conclusions like the one above.

In a free market for credit, interest rates are set just like other prices: by supply and demand. Assuming constant demand, interest rates will be low when lendable capital is abundant (due to increased saving) and high when it's scarce. Under these conditions, interest rates can never be too low; the benefits of abundant capital are diminished only by the fact that holders of capital must accept smaller returns than they would if it were more scarce.

The idea of an interest rate that is too low becomes meaningful only in a credit market governed by a central bank. Central banks lower interest rates not by increasing the supply of real capital, but simply by minting new money and lending it out. Since these loans aren't backed by real savings, they tend to have an inflationary effect. The problem is not that interest rates are too low by some absolute standard; it's that rates are lower than is justified by the availability of capital.

And therein lies the problem. Deficit spending does have the effect of raising long-term interest rates, but it does so by diverting capital away from the private sector. This does nothing to mitigate the problems caused by artificial depression of short-term rates by the central bank, and---I'm going out on a limb now---may even aggravate them. Since deficit spending shrinks the pool of available capital, I suspect that it should widen the gap between the natural market rate of interest and the artificially depressed interest rates imposed by the central bank. As I said above, it's this gap, and not the fact that interest rates are low in absolute terms, that is the real problem.

Share this

The idea of an interest rate

The idea of an interest rate that is too low becomes meaningful only in a credit market governed by a central bank.

Well, yes. Our credit market is governed by a central bank. As is every credit market in the world, I believe. Thus, it's quite meaningful to speak of an interest rate that is too high or too low. However:

The problem is not that interest rates are too low by some absolute standard; it’s that rates are lower than is justified by the availability of capital.

That's incorrect. Interest rates in the presently existing credit markets can be judged too high or low by some absolute standard, and cannot be judged against what would be justified by the availability of capital - because there is no free credit market within which such an assessment can be made. Interest rates are only meaningful within the context of a particular monetary regime, whether that money is gold, fiat credit, or free credit. Neither you nor anyone else knows how much capital would be available or demanded right now under a free credit regime, because right now there is no free credit regime.

So how can the interest rate be assessed as too high or too low? Only by looking at some desired outcome of monetary policy - low inflation, low unemployment, high growth, presidential re-election, whatever. What those goals are, or should be, and whether the interest rate is too high, too low, or just right to meet those goals, and whether those goals are even possible to attain through monetary policy... all this is subject to extensive debate bordering on the absurd. Furthermore, what is too high for you may be too low for me. But it's still an absolute standard - I absolutely say it's too low, and you absolutely say it's too high, and if we happen to be Fed governors we get to vote on it.

Deficit spending does have

Deficit spending does have the effect of raising long-term interest rates, but it does so by diverting capital away from the private sector.

Maybe. It depends on what you're comparing it to.

Government spending financed by borrowing versus government non-spending does indeed divert productive resources away from the private sector. But then so does government spending financed by taxes rather than borrowing. And neither of these necessarily affects interest rates, at least not directly; they just put more of the economy in the inefficient hands of the government.

Government spending financed by borrowing versus the same government spending financed by taxes might divert capital away from the private sector, but it might not. It's not clear. Ricardian Equivalence says that it doesn't make a difference. But on the other hand, there are some reasonable counterarguments to Ricardian Equivalence. But on the third hand, empirical evidence suggests that deficits don't affect interest rates.

In my uninformed opinion, Ricardian Equivalence is broadly true but not perfectly true. Enough people behave as normal persons rather than rational optimizers that there is some crowding out effect, but the effect is small enough to not be measurable against the normal noise of economic activity.

“Assuming constant demand,

“Assuming constant demand, interest rates will be high when lendable capital is abundant (due to increased saving) and low when it’s scarce.”

Umm… No. It is the reverse. Assuming constant demand, when capital is abundant the price of that capital falls – that price is the rate of interest it demands. I think that what you are trying to argue is that in order to place ever increasing amounts of debt in the hands of debt holders (lenders of capital) interest rates must rise to attract that demand. In effect demand does not remain constant because you use the change in price to increase demand.

What we have here, and what really explains the lack of more dramatic increases in interest rates on the fiscal deficit, is an abundance of capital available to purchase government debt at rather low yields.

This capital comes primarily from four sources:

1. Recycled “petrodollars” that flow from consumers pockets to oil producers and then come back and are parked in govt securities - it is, perversely a kind of tax that funds the govt but only via a circuitous route.

2. Recycled current account surpluses that other nations have with us (e.g. China, Japan, etc) the funds from which return here because we have one of the few true reserve currency (for now) and in a “mercantilistic” jobs creation/inflation abatement program in China that requires they continue to buy dollars almost without end.

3. Generalized surplus savings among companies and foreign individuals which eventually find safe haven in this country and in our currency

4. Seniorage: the world wants to hold and use dollars and we get a rather attractive zero interest loan from it.

Eddie: I don't have to know

I don't have to know what the interest rate would be in a free market. And I don't have to know how much deficit spending affects interest rates. My point is that if interest rates are too low, then the increase in interest rates caused by deficit spending (if any) does not have the same salutary effect as raising the federal funds rate (i.e., abstaining from inflating), and may even have the opposite effect.

Why do you think that Ricardian equivalence is broadly true? Do you know anyone who bases his savings decisions on predictions about future tax rates?

Garth: Thanks for catching


Thanks for catching that. It was just a typo---which I hope is evident from the rest of my argument---and it's fixed now.

Also, remember that demand is a function, not a scalar. When I said "assuming constant demand," I meant to assume a constant relationship between interest rates and the quantity of money people are willing to borrow at that rate, not to assume that demand for loans is perfectly inelastic with respect to the interest rate.

Finally, could you elaborate on the part about recycled petrodollars? Are you saying that the oil industry holds an unusually high proportion of its assets (compared to, say, the auto industry) in government securities? Why?

Indeed. There are no hard

Indeed. There are no hard assets in the Saudi Kingdom (or most oil producing countries aside from, say, the U.K.), Russia, etc etc etc and so on and so forth. Look at any graph of the growth of international reserves of any of these countries, then take a look at the dollar-denominated assets of the wealthiest benefactors (private or public) from the oil windfall and you will see a massive accumulation of dollar assets with the vast majority parked in US Govt Bonds or Mortgage-backed securities (Fannie Mae etc).

Even if it is not the majority of the inflows, it's not far off and simply swamps most other single sources of inflows into the public securities market.

No oil producing country would ever "convert" it's received dollars into local cash from 1) the massive increase in the money supply and subsequent inflation; 2) the attendant "Dutch Disease" effects; and 3) their own currencies are not real money and they know it.......

BTW when you say "oil industry" I hope you recognize that the vast majority of output comes from non-US-corporate entities and mostly from state-or-state-controlled producers.

Thanks for the commentary on

Thanks for the commentary on dollar/bond cycle we seem to have with various large trading partners. I happen to agree with it, though I've seen others argue that these are small compared to our total trade (not believable) or small compared to our (US) total bond market (true as far as it goes).

I'd like to go back to what a too low or high interest rate is, and how a large deficit even at low rates might harm the economy. I'd like to put forward that there is an interest rate and a deficit size that is efficient in that it produces a larger long term growth rate of the economy. We may not know what those parameters are precisely, but if the government is going to control rates (and deficits?) properly it must try to estimate them. I guess I'm of the rising tide theory of government.

There are arguments against allowing high inflation because it harms long term growth, and arguments for running deficits that "invest" in various aspects of the country (education, roads, defense) to increase long term growth. I'd prefer to focus not on their values, but large fluctuations.

I suspect that most consumers (the wealthy more so) could adapt to fairly different levels of "investment" or inflation with minimal impact on longterm growth. The extremes of Sweden/HK, and Brazil/Switzerland come to mind as each has actually performed rather well through these extremes. What people really don't like is uncertainty (perhaps irrationally so?). Rapid changes in public "investment", inflation, (long term?) interest rates, and trade deficit limit our ability to plan and make good investments. They can invalidate recently made and seemingly appropriate purchases/sales on either side of the transaction. Just as both sides of an "efficient" transaction may benefit, exogenous effects can make both sides lose even after the fact.

My real fear in running large deficits is not their short term effect on rates. I don't believe that there's a great benefit to cutting today's taxes, only to raise them later to pay off today's deficit, but there may be some utility to be made in the float as GNP rises. I've got a 2% car loan that I could have paid cash for, but didn't because I think inflation will stay above 2% for the life of the loan. I'm not going to retire on the gain, but the problem comes when you can't handle the call.

The current budget is ~$2300B, deficit ~$450B, and trade deficits to China, Japan, and oil nations ~$300B with almost all of that coming back into bonds. Spending is also disproportionately dependent on home building, and mortgage refis due to low interest rates. My point is that there isn't significant slack. We have to borrow as much tomorrow as we did today, and if we were cut off even marginally it could have uncontrollable effects on long bond rates (private and public), and on domestic consumption. If even 5% more capitol had to be raised internally over the course of a year, government supply would be inelastic, and private debt supply all too elastic. We're borrowing lunch money from a crack dealer, the worst of all worlds.

Such a rapid and large shift in interest rates would cause what I would call inefficient future investments (or lack of the alternative), and make seemingly sound previous investments unviable. So the real reason not to run giant unsustainable deficits (budget/account) even at very low interest rates, is not just that you have to pay them off (in inflated dollars), but that investors might not be able to borrow except at outrageously high rates, when GNP growth has become dependent on low rates. Exports and imports will come into balance eventually, but over a significantly longer term especially since any investment in capitol or labor productivity will be very very expensive. Note that our foreign competitors will have depreciated factories and trained workforces. To avoid that catastrophe the government needs to manage the knobs it has wisely and avoid the free lunch psychology.

Or perhaps not. It could be that we are irrationally adverse to possible global economic crisis, and that the years of higher growth make up for the eventual stagflation. I'll hedge with foreign denominated bonds, just in case.

I would add a couple of

I would add a couple of quick observations on all of this.

The first is that while the evidence surrounding the direct relationship between fiscal deficits and interest rates is flimsy, there have been more robust studies about Ricardian Equivalence and the long run impact of deficits on savings. This latter should prove more recessionary when the crows come home to roost than any direct impact of rates. And I would add, at some point we do have to bring the fiscal back into line... or inflate it away.*

The second, and even more important point, is that prices are determined on the margin: I worked on a bond trading floor for ten years and I can vouch that the entire stock of outstanding bonds is priced on the latest settled trade. To whit: today we see interest rates at x, but the moment demand for our debts (either on concerns over fiscal deficits, or the price of the dollar or whathaveyou) diminishes by even a small amount, that price can move fairly quickly -- unless there is some form of intervention which is possible but would be a signal to the market to sell even harder making such an intervention worse then useless.

*One word on the negative impacts of inflation. It hurts the poor disproportionately since they typically do not have financial tools to cope with it, do not have wages that adjust quickly, and otherwise lag price rises more than those more well to do.... Having said this, the more important economic damage done by high inflation is when it is unpredictable (and this certainly is the case during the transition from low to high inflation) investment decisions are typically interrupted. It is difficult to plan over any horizon when prices are changing in an unpredictable fashion. It also can severely damage industries for whom products have elastic demand curves. I would also note that while a shift to higher inflation is a boon to debtors (assuming fixed rate obligations) it is bane to savers, especially in the short term though in the long run most financial assets act as though indexed.

Oh, and to the point of the

Oh, and to the point of the price of bonds being determined by the last trade: the inflows we have seen from recycled pertodollars/foreign current account surps/global savings have been propping up the bond market on the margin. We don't recycle the stock every week, but at each auction there are plenty of buyers supporting the price. The secondary market flow, while important, is not deterministic on this front.

To continue on Garth's last

To continue on Garth's last comment, but taking it a different route, the U.S. has reached an interested paradox in the bond market. Regardless of whether it is petro-dollars or currency strategy, foreigners are buying our debt without much thought about price. Even if you wanted to invest in every other treasury market besides the U.S., they are not as big or liquid as ours. This makes our Fed's goal of slowing down inflation very tough because long rates have not moved in synch with short rates. Unless foreigner demand slows down/dries up or they start caring about price, one way to contain inflation will be to borrow MORE on the long end to increase supply and drive up yields. What politician wouldn't love that?

I like your idea Patinator.

I like your idea Patinator. It would be better to cut spending or increase consumption taxes to reduce the deficit, but given the choice I'd much rather be lending out low long bonds that short ones. The fed has been issuing much more debt as short bonds in the last 4 years (with the demise of the 30yr in 2001). With the yeild curve so flat it would make a lot of sense to switch back, and that may come to pass.
Old news, but if we are going to inflate our way out of this mess, I'd much rather have 30 years to do it rather than 5. However, if Garth is right, even a twitch could have Jenga tower consequences.

p.s. actually I've been looking for data on the average length of public securities, if you know where recent data (>2003) can be found on the FRB website or elsewhere. How much shorter than 60mo can we go?

Kurth, Here is the closest


Here is the closest thing that I could find. (page #8)

I was not able to find this data historically from this site, but you could send them an email. However, as you suspect the average maturity issued and outstanding has been dropping. When they open the 30 year back up in January, it should help although I heard they are only planning only ~$10B for that tenor.

I think I found the raw data, but you will need to download all issuences then net out repurchases from the following website:

I hope your database/excel skills are good.