Matt Yglesias has a post advocating fiscal stimulus in the form of checks sent to all Americans. While discussing the bonds that would be issued to finance the increased deficit, Yglesias makes this claim:
The federal government needs to pay interest on the bonds, but that interest becomes Fed profits which are rebated to the Treasury so there’s no actual cost.
Of course, the total federal debt will go up, but the more important “debt held by the public” will not since the extra debt will, by law, be perpetually held by the Fed rather than by the public.
That would be true if the bonds were bought with zero interest currency notes, and the currency notes stayed in circulation forever. Most of the debt, however, will be purchased with interest-bearing bank reserves. This is because a large permanent increase in the stock of zero interest currency would likely cause the Fed to overshoot its 2% inflation target. Maybe not right away, but certainly after the economy recovered and interest rates rose above zero.
So let’s assume that the deficit is financed by issuing Treasury debt, and the Fed buys the debt with interest-bearing bank reserves. In that case, there is not likely to be much Fed profit to offset the interest burden on the Treasury. Yes, the Fed will earn interest on the bonds it purchases, but it will pay interest on the bank reserves that it injects into the economy.
In general, the interest rate on bank reserves is roughly equal to the interest rate on T-bills. While it is usually (but not always) the case that interest rate on bank reserves is below the interest rate on longer-term bonds, that sort of “profit” could be earned simply by shortening the maturity of the Treasury’s outstanding debt. For various reasons, the Treasury prefers to borrow by issuing bonds of a wide range of maturities. If the Fed bought longer-term bonds with bank reserves they would probably earn a profit, but there is risk involved.
It’s also possible that interest rates will stay at zero forever. But in that case there would be no cost in financing the deficit even if the debt were not purchased by the Fed. So money creation does not perform any fiscal miracles.
On the other hand, because the demand for currency rises over time, the Fed earns a certain amount of seignorage from adding to the stock of currency. But that profit (sometimes referred to as the “inflation tax”), occurs regardless of what fiscal policy is doing. And it’s a very small share of GDP, relative to the entire federal budget.
PS. Some might argue that the US should copy Japan’s “zero interest rates forever” policy. But Japan did not achieve that outcome with easy money; they did so by producing ultra-slow NGDP growth, which drove the Japanese natural rate of interest down to zero. People who favor proposals to monetize debt and/or permanently hold interest rates at zero, are generally people who favor expansionary policies regarding aggregate demand. Anyone in that camp should NOT be citing Japan, which among the major economies has the worst performing aggregate demand growth in modern history. That’s why Japanese interest rates are stuck at zero. Nonetheless, I recall seeing a few MMTers cite Japan as an example of a zero interest rate policy.
READER COMMENTS
Lord Canes
Dec 15 2020 at 8:00pm
The example of Japan is far more complicated than usually understood, and presented here.
During the decades usually cited as Japanese stagnation, average hours worked per year per worker were radically reduced. From 1980 to 2011, average hours worked per worker in Japan fell by 19%. (the FRED series halts at 2011). We are talking about four hundred additional leisure hours per year, or an additional 10 weeks if one considers a week to be 40 hours of work.
Until the Wuhan virus pandemic, the number of job openings in Japan was vastly greater (50% more) than the number of job hunters, obviously a huge success story
Though stereotypes remain, Japanese presently work less hours per year than Americans and far less than Russians . The Japanese have to go a long ways before they catch up with the Germans in leisure time, however.
https://stats.oecd.org/Index.aspx?DataSetCode=ANHRS
Whether for better or worse, and I would argue for worse, Japan offshored a tremendous amount of industry in the lost decades (a strong yen and good wages in Japan), and this had some effect reported national income or GDP.
That is an interesting proposition of Scott Sumner’s that the Fed paying interest on excess reserves prevents inflation and simultaneously undercuts the proposition that money-financed fiscal programs are possible or successful.
The Bank of Japan pays 0.1% interest, on only some reserves, and yet Japan has lower inflation than the US, as measured (presently Japan is in deflation).
In sum, the modern macroeconomics profession has yet to resolve what is really happening with quantitative easing and fiscal deficits.
Leading lights of the profession are often (usually?) in diametrically opposed positions.
Scott Sumner
Dec 15 2020 at 8:15pm
You said:
“The example of Japan is far more complicated than usually understood, and presented here.”
I was not discussing “the example of Japan”, I was discussing the example of aggregate demand in Japan. That’s why I did not discuss all the issues you raised, they have little or no bearing on my point. (Even if you look at aggregate demand per member of the labor force, Japan still has had appalling low growth in AD.)
There’s no mystery as to why interest rates are so low in Japan; that’s what you’d expect when you have more than 25 years of near zero NGDP growth.
Lord Canes
Dec 16 2020 at 3:17am
There’s no mystery as to why interest rates are so low in Japan; that’s what you’d expect when you have more than 25 years of near zero NGDP growth.–Scott Sumner
Interesting question: If interest rates are low in Japan, why do not Japan lenders lend offshore, until rates in Japan equalized with the global level?
I happen to believe more stimulus is needed across the developed world, fiscal-monetary, and in Japan.
That said, I found an even better table at FRED.
https://fred.stlouisfed.org/series/AVHWPEJPA065NRUG
Of course, leisure time is not reported in GDP figures. If leisure time was given a value, then NGDP would be much higher than reported in Japan.
Scott Sumner
Dec 16 2020 at 4:18pm
The interest parity condition explains why Japan has lower interest rates. Its currency is expected to appreciate over time.
And leisure is not a part of aggregate demand because it does not cost money. It is a part of living standards, but that’s a separate issue.
And again, AD growth has been really slow even if you account for leisure.
Mark Barbieri
Dec 15 2020 at 9:22pm
I hope that he was advocating for astronomical checks. If the checks won’t cost the government anything, let’s not stop with a measly $1,200/person. Wouldn’t $120,000/person be much better?
jj
Dec 16 2020 at 12:18pm
I’ll take $120 million, thanks. Might as well send blank checks in case anybody wants more than that.
David C Everson
Dec 17 2020 at 1:48pm
And let’s crank up the minimum wage some more while we are at it. If $15 is good $30 will be twice as good.
Market Fiscalist
Dec 15 2020 at 10:57pm
An MMTer (who understood the idea of NGDPT) might respond as follows:
‘If NGDP is below target then this can be addressed by increasing the government deficit (for MMT this means increasing the total quantity of govt bonds + money held by the public). This can be done by either increasing spending or decreasing taxes.
If at a later date NGDP rises above target then this can be addressed by decreasing the government deficit. This can be done by either cutting spending or increasing taxes.
As long as both parts are done fairly then the costs of addressing the initial fall in NGDP is a free lunch since the later action is simply reducing the the total quantity of govt bonds + money held by the public to its original level and there are only winners (those who benefit by avoiding a fall in NGDP) and no losers.
Overall interest rates may vary over the period and cause some people to benefit and others to lose but that is small beer big-picture wise.
People who favor monetary policy over fiscal policy may object that they could do the job just as well (or better) but is this really true when recent episodes seem to have shown that reducing interest close to zero seems not to have made much of a dent on NGDP trends?’
(remember I’m being an MMTer here !)
Jerry Brown
Dec 15 2020 at 11:32pm
Yes that is pretty good. An MMTer might consider government bonds to be almost the same as the monetary base for most purposes also.
If NGDP is below target MMT would say there needs to be more spending. Could be private sector spending achieved by reducing taxes or could be increased government spending without raising taxes. And in either case the bond issuance that the law says must go with deficit spending has little to no effect on anything important.
Scott Sumner
Dec 16 2020 at 4:22pm
A couple points:
Obviously deficit spending did not work in Japan, as their massive deficits during the 1990s and 2000s were accompanied by deflation. When Abe cut the deficit sharply after 2013, NGDP actually rose.
Even if that policy worked (and I don’t think it would) it doesn’t really have any bearing on Yglesias’s claim, which is incorrect as he ignores the interest paid on bank reserves.
Market Fiscalist
Dec 16 2020 at 9:29pm
I agree that my comment was a bit tangential to your critique of Yglesias’s point. I was really just thinking aloud as (prompted by your recent posts) I’d been trying to understand MMT a bit better.
Why do you think that fiscal expansion in Japan failed? Was it Ricardian equivalence or something else ?
Scott Sumner
Dec 18 2020 at 1:23pm
Fiscal stimulus is simply not very strong. It could be a mixture of Ricardian equivalence and monetary offset, but since the effect of fiscal stimulus is small, you don’t need much offset.
James
Dec 15 2020 at 11:22pm
“That would be true if…”
No. If the government uses the money to buy goods and services, those goods and services (or the resources used to produce them) are no longer available for other uses. That is the cost of government spending and no financing strategy can change that.
Sean Kemsley
Dec 16 2020 at 12:35am
I think that’s a great point, I like thinking in those terms. I think it’s especially relevant if you believe, like me, that government will use those resources less efficiently than the private sector. However, just playing devil’s advocate here, if the economy’s resources are being underutilized due to a slump in aggregate demand and a sluggish transition to a new equilibrium price level (sticky wages and contracts or whatever cause), then is it not the case that the government spending could make use of certain resources that otherwise would have been left idle absent the government intervention?
James
Dec 16 2020 at 11:50am
There is no government program where someone is making sure that only excess capacity is being used. If the government spends money on building an infrastructure project such as a bridge, that requires land, workers, equipment, energy and materials. Does anyone really believe that all of the inputs to the project are excess capacity?
Even if (theoretically) the government uses borrowed money to pay unemployed people to operate rides at an abandoned amusement park, this uses non-slack resources too. It takes energy to power the machinery. It also uses the time that the unemployed could have used to look for jobs producing goods and services that people actually want.
Maybe someone could imagine a form of public spending that would only use idle resources but actual government spending programs do not work in that way at all.
Henri Hein
Dec 16 2020 at 3:00pm
This is a point about Keynesian theory that has always confused me. Even during a recession, most resources in the economy are still in use. If government starts an arbitrary project using arbitrary resources, it is extremely unlikely to hit on just the unused ones. I’m pretty sure no economist considers all resources fungible. So what is the Keynesian story for how stimulus projects would activate idle resources?
Jose Pablo
Dec 16 2020 at 10:53pm
You don’t need to use “only” excess capacity. The intervention would make sense if the % of excess capacity used is bigger than the difference in efficiency between the public and private investments. In this case the opportunity cost of using “non excess” capacity would be more than offset by using “some” excess capacity with zero opportunity cost.
The probability of this happening is bigger the bigger the excess capacity and the lower the “efficiency gap” between the public and private investments. Which could be (theoretically) fine-tuned by choosing “when” and “where” using this policy.
Which I find extremely difficult to believe is that the government will know when to stop running the empty amusement park. It will run until the last public worker there dies.
Henri Hein
Dec 17 2020 at 12:35am
What you are saying makes sense, but it seems to go against Keynesian theory. At least in the original incarnation, the project doesn’t matter. If we use stimulus to prepare for an alien invasion that never happens, to use Krugman’s famous example, efficiency is pretty much zero.
Jose Pablo
Dec 17 2020 at 10:07am
I was not trying to be Keynesian (I am not). I don’t believe the government has the smallest chance of get and properly aggregated the amount of information dispersed among the economic agents required to get right the amount of excess capacity at any given time or to tell the sensible projects from the useless ones. And this judgements matter.
But more than anything else I don’t think the government is well equipped to include “time” in the equation. Excess capacity is a function of time with a too steep first derivative for the bureaucrats to grasp it, taking into account their incentives scheme.
In the biggest Keynesian experiment of all times, the Roosevelt’s New Deal, we (as Professor Walter Williams (RIP) used to point out), very likely traded a few years (maybe less) of reduced excess capacity in exchange for a much poorer, maybe an order of magnitude poorer, society nowadays (and could have been even worse absent WW II).
It turned out that in the long term, some of us are still very much alive.
James
Dec 17 2020 at 12:24pm
I was only thinking in terms of the idea that government spending could, in some scenario, be free. It cannot. If the government uses any resources that had an alternative use in the private sector, those resources are unavailable to the private sector.
Matthias
Dec 16 2020 at 2:19am
Yes. Though the typical argument you see advocating for this kind of programme also includes a remark saying that the economy is operating below capacity.
Scott Sumner
Dec 16 2020 at 4:24pm
Yes, you are correct. If goods and services are produced there is an opportunity cost. I was merely referring to the budget cost.
Thomas Hutcheson
Dec 16 2020 at 6:21am
Yes he is wrong on this, but let’s cut him a smidgen of slack in that his main argument is that other kinds of relief are politically difficult (although I do not know why he thinks Fed checks would be easier). Conceptually, the larger error is in not having a theory of how this would play with Fed policy. Would it lead the Fed to get inflation up to an NGDP targeting level (>long term target) and actually have some “simulative” effect as he apparently believes or, more likely, would the Fed just reduce other bond purchases in pursuit of whatever their target is (which does NOT appear to be stable prices and maximum employment).
Rajat
Dec 17 2020 at 4:56am
Great clear post, Scott. Now the obvious question. Yglesias seems to have a strongish following amongst the ‘new’ left-neoliberal set. He half-gets (macro)economics but he’s not an economist (nor claims to be). When are you putting your moneyillusion blog on Substack? Maybe you can afford Orange County after all 😉
Rajat
Dec 17 2020 at 6:48am
Oops, I meant the Hollywood Hills.
Scott Sumner
Dec 18 2020 at 1:29pm
I don’t have any plans to go onto Substack, but if I thought I could make even 10% as much money as Yglesias I would probably do so. I doubt whether many people would pay for TheMoneyIllusion.
If I read this correctly, he has just below 10,000 subscribers, each paying $8/month. (I’m one of them.)
https://twitter.com/mattyglesias/status/1339250351237689348/photo/1
Comments are closed.