Scott Sumner  

Does QE reduce the public debt burden?

PRINT
Trump's Trade Fallacies... Pluralism and Civil Society...

Bill directed me to an article that suggests the answer is yes:

Japan for years has been renowned for having the world's largest government debt load. No longer.

That's if you consider how the effective public borrowing burden is plunging -- by one estimate as much as the equivalent of 15 percentage points of gross domestic product a year, putting it on track toward a more manageable level.

Accounting for the Bank of Japan's unprecedented government bond buying from private investors, which some economists call "monetization" of the debt, alters the picture. Though the bond liabilities remain on the government's balance sheet, because they aren't held by the private sector any more they're effectively irrelevant, according to a number of analysts looking at the shift.

"Japan is the country where public debt in private hands is falling the fastest anywhere," said Martin Schulz, a senior economist at Fujitsu Research Institute in Tokyo.

While Japan's estimated gross government debt is now over twice the size of the economy, according to Schulz's calculations using BOJ data, the shuffle of holdings from private actors like banks and households to the central bank is having a big impact. It means debt in private hands will fall to about 100 percent of GDP in two to three years, from 177 percent just before Prime Minister Shinzo Abe took power in late 2012, he estimates.


It's possible to make a good argument that Japan's debt burden is falling, but I don't believe this is because the BOJ holds lots of Japanese Government Bonds (JGBs). Rather they are benefiting from very low interest rates, and a strong likelihood that rates will stay low for the foreseeable future.

Consider four possible cases:

1. Japanese interest rates stay low indefinitely.

2. Japanese interest rate eventually rise above zero, and the BOJ continues to pay market interest rates on bank reserves.

3. Japanese interest rates eventually rise above zero, and the BOJ stops paying IOR. They sell off most of their stock of debt to prevent hyperinflation.

4. Japanese interest rates eventually rise above zero, and the BOJ stops paying IOR. They hold on to their stocks of JGBs bonds, leading to hyperinflation.

In case 1, the Japanese government benefits from the near zero rates, but they'd benefit regardless of whether the BOJ held the bonds, or whether the private sector held the bonds. It's the near zero rates that benefit the Japanese government, not the fact that QE occurred.

In cases 2 and 3 it may seem that BOJ debt monetization offers a way to reduce the debt burden. But I think that's misleading. In case 2, the BOJ must continue to pay interest on the reserves that are backed by the debt it has purchased. Since the BOJ is part of the Japanese government, that's still a debt burden. In case 3 the stock of bonds is eventually sold off, so the burden of the debt is still there in the long run.

Case 4 seems to offers the greatest prospect for reducing the burden on the debt. If market rates rise above zero, however, and IOR stays at zero, then the BOJ must sell off its huge stock of debt or face hyperinflation. Obviously if they hold onto the JGBs and there is hyperinflation then the burden of the debt falls sharply. But there is no indication that the BOJ plans to allow hyperinflation in the foreseeable future. So I view case 4 as very unlikely.

Bottom line: If something seems too good to be true, it usually is. That's not to say governments are not seeing any benefits from low rates. If interest rates average only 1% over the next 100 years (which doesn't seem that much of a stretch for a country like Japan), then even a debt of 200% of GDP implies an annual interest cost of only 2% of GDP.

Countries with their own currencies, such as the US, Britain, and Japan, are facing pretty modest debt burdens at current interest rates, and indeed the same is even true for countries without their own currency, but with a credible commitment to service their debt, such as Germany. Even Greece's debt burden would be easily manageable at German, Japanese or American interest rates. If Greece could get Germany, Japan or the US to co-sign their government loans, then the Greek government could easily service its debt. However, if they actually did find someone gullible enough to co-sign their loans, then Greece would default on the debt, as Greece is especially susceptible to the moral hazard and time inconsistency problems.

Greece has in fact been in default on its external financing for fully 50% of its time as an independent country.

Comments and Sharing






COMMENTS (26 to date)
nick writes:

Professor Sumner- Is there any reason to think that the market expectation for which of the above scenarios is followed has an impact on the real inflationary effect of QE? Put differently, I've often wondered whether our (probably correct) belief that the Fed will eventually sell their balance sheet back to the public once inflation appears is a primary reason why QE doesn't lead to inflation in the first place.

Yaakov writes:

Will the BOJ have a problem selling off the debt when the interest rates rise? Wont it have to sell at a large discount, incurring large losses?

Market Fiscalist writes:

Isn't there a fifth option?

5. Japanese interest rates eventually rise above zero, and the BOJ stops paying IOR. They run govt surpluses to prevent inflation and use the surplus to reduce the debt.

What your case 2 shows is that governments never really need to sell (or buy) bonds at all, they could do everything with IOR - but as in both cases they have to pay out the same amount of interest, and this interest ultimately has to be paid for either by taxation or inflation - it is largely irrelevant which they use.


Scott Sumner writes:

Nick, Yes, only QE that is expected to be permanent has a significant impact.

Yaakov, Hard to say, the EMH says that on average the BOJ would neither gain or lose, but there is some risk involved. Just another reason why central banks might want to steer clear of the zero bound.

Keep in mind that the BOJ is part of the Japanese government, and any capital loss to the BOJ reflects an equal gain to the Japanese Treasury, so it nets out to zero.

Market, No, at positive rates and no IOR, even fiscal surpluses would not be enough to prevent hyperinflation. There'd still be a huge mismatch between base money supply and demand.

Market Fiscalist writes:

'at positive rates and no IOR, even fiscal surpluses would not be enough to prevent hyperinflation'

Surely there would be some level of taxation that would allow you to reduce NGDP to whatever level you wanted it to be , no matter how great the initial money supply ?

baconbacon writes:
Surely there would be some level of taxation that would allow you to reduce NGDP to whatever level you wanted it to be , no matter how great the initial money supply ?

Nope. But they could reduce RGDP (almost) as low as they want.

Market Fiscalist writes:

baconbacon,

If I have learned one thing from reading Scott's blog it is that if you control the money supply you control NGDP. There is no more direct way to control the money supply that using govt deficit/surplus to do so.

I therefore conclude that taxation could always be used to control the money supply and hit an NGDP target.

Whether it would be practically wise - I doubt.

Matthew Waters writes:

Using surpluses to pay down debt while BOJ's balance sheet remains constant can't get around neutrality of money. The amount of Yen out there stays the same. It's going to be spent on something outside the zero-lower bound.

But the BOJ would have to do MORE QE for monetary base to remain constant. They would have to do QE in the midst of hyper-inflation. The bonds held by the BOJ will naturally mature and the principle is effectively destroyed money if the BOJ does not buy more bonds.

"Unwinding balance sheet" would be a mix of bonds maturing and selling the bonds. I think Yellen has hinted there's no reason reason to have to sell the bonds, except for the traditional Fed Funds rate mechanism. Sell enough bonds to raise rates if necessary, then let the rest mature.

baconbacon writes:
If I have learned one thing from reading Scott's blog it is that if you control the money supply you control NGDP.

Perhaps you should ask for your money back?

Even in SS's world it is the CB that controls the MS, not the tax authority. If the TA tried the reduce the MS via taxation they would always be overwhelmed by a CB that was trying to inflate.

In the real world markets control NGDP by setting the exchange rate between money and goods and services. Say you attempt a 100% sales to be paid by the seller of the good. Who would sell any good and then remit 100% of the proceeds to the government? If they reduce it to 99% then goods and services must be priced at 100x their pretax price for the same nominal profit. If you try to tax from the other direction, say an income tax, then employers have to pay 100x the salary to pass on the same nominal wages.

High tax rates make money less valuable, not more so, and eventually governments get to rates where alternate means of trade (foreign exchange, barter) are more profitable that in local currency and so lose any control whatsoever. In the late stages of hyperinflation you often 9always?) see price level increases outstripping MS increases by factors of 100 or 1000. At points in the Weimar hyperinflation bankers argued that there was a shortage of money since the amount of money in circulation was no longer sufficient to pay for all the goods and services that exchanged hands.

Governments routinely hit limits of tax collection well below these extreme examples, for a variety of reasons which make practical curtailing of an out of control MS unlikely on their own.

Market Fiscalist writes:

Undoubtedly Scott would say that monetary offset would prevent the fiscal authority from controlling the money supply - I am assuming that away. I was envisaging that the fiscal authority just creates/destroys money by running deficits/surpluses (no bonds).

baconbacon writes:

@Market Fiscalist,

At what level of taxation would you stop accepting US dollars as payment (be it wages or sale of goods you produced)?

Matthew Waters writes:

"Undoubtedly Scott would say that monetary offset would prevent the fiscal authority from controlling the money supply - I am assuming that away. I was envisaging that the fiscal authority just creates/destroys money by running deficits/surpluses (no bonds)."

I think that's the misunderstanding. Deficits and surpluses do not destroy and create money. The central bank does that.

You're assuming that as the government runs surpluses, the central bank does not replace bonds as they mature. The government would take tax dollars and pay the principal. When the central bank decides not to reinvest the principal, then the central bank is destroying money.

The CB also destroys money by selling bonds on the open market.

Market Fiscalist writes:

@Mathew

'I think that's the misunderstanding. Deficits and surpluses do not destroy and create money. The central bank does that.'

There is no particular reason why money could not be created/destroyed like that - its just that the current monetary system has chosen to do it via OMO. However even today the CB/govt contrive to produce inflation each year by the CB buying back a proportion of govt debt - this is helicopter money in disguise and earns the govt an "inflation tax".

bill writes:

I continue to digest that original article along with this post. One thing that strikes me is the supreme irony of the Japanese situation. My whole life I've been told and thought that a highly indebted government might just inflate its debt away. The Japanese seem to have done the opposite.

Matthew Waters writes:

@Market

Yes, there's no reason the BOJ couldn't create helicopter money or eventually monetize the debt. That was case 4 in Scott's post.

But saying "the BOJ could let hyperinflation happen" is like saying "Obama could invade Canada to steal Maple Syrup." That is a thing that Obama could do if he really wanted to.

But simply buying up the debt does not ensure it will be monetized. The bank would have to continue loose policy even as inflation increased. There's no indication that would happen.

Market Fiscalist writes:

@Mathew

I'm actually saying the opposite - that the Japanese could avoid hyperinflation without either IOR or bond sales by hoovering up excess money via a sufficiently large budget surplus.

As long as V increased slowly over a number of years there is no reason this would not be viable. Might not be optimal though.

Lorenzo from Oz writes:

Market Fiscalist: "There is no more direct way to control the money supply that using govt deficit/surplus to do so." Apart from, well, printing money.

Market Fiscalist writes:

@Lorezno

OK, but (assuming you're the govt) once you've printed it , what's the accounting going to look like as you distribute the money that won't appear either as an increase in the CB balance sheet, or a decrease in the govt deficit ?

Scott Sumner writes:

Market, Budget surpluses don't affect the money supply. They boost money demand, but not enough to prevent hyperinflation.

If the surpluses were somehow used to reduce the money supply, you'd be back to case #3.

bill writes:

The accounting would just result in a negative equity account.

Richard Solomon writes:

[Comment removed pending confirmation of email address. Email the webmaster@econlib.org to request restoring this comment. A valid email address is required to post comments on EconLog and EconTalk. Broken url removed.--Econlib Ed.]

Benjamin Cole writes:

Interesting post.

I wonder if this is more fodder for the Adair Turner-Michael Woodfords.

Forget about QE. IOR. Reverse repos. Who understands what is happening anyway?

Just do helicopter drops. That is, money-fnanced fiscal programs. I would prefer tax cuts. Just cut taxes and print money to make up for the lost revenues.

The argument each year become simple: How big are the chopper drops?

Market Fiscalist writes:

'Budget surpluses don't affect the money supply'

If the govt runs a surplus and neither they or the CB buy or sell any bonds - what is the difference between that and a reduction in the money supply ?

Scott Sumner writes:

Market, I don't follow your question. If the Treasury runs a surplus, they presumably either fail to roll over bonds that mature, or even buy back existing bonds. Neither action affects the money supply.

Market Fiscalist writes:

I was thinking of a very simple model where the govt either runs a deficit which it funds by printing money, or runs a surplus which leads to it accumulating money which it destroys.

This would be functionally the same (I think) as deficits being covered by matching bonds sales by the govt and purchases by the CB and vice versa for surpluses.

Richard Solomon writes:

Dear Professor Sumner,

Bloomberg’s quote from Martin Schulz is somewhat incomplete. Martin actually believes the BOJ’s transfer of government debt from private hands to public sector accounts is an indication that the great claw-back of private sector wealth has begun in earnest. Beset by a declining population and workforce as well an aging population, Japan’s pension Ponzi scheme will collapse without a claw-back of wealth held by the elder generation. As per your scenario number 4, he believes the government will likely have no choice other than to inflate government debt away. You can read the original source of Bloomberg quote here: http://beaconreports.net/whats-next-abenomics/

Comments for this entry have been closed
Return to top