Scott Sumner  

Money is fungible

PRINT
The Center Will Hold: I'll Bet... Come Away With Me...

Yesterday I discussed one bizarre idea that has popped up in the wake of the Great Recession. Today I'll look at another. The newly elected leader of Britain's Labour Party has proposed a "People's QE". The plan is discussed in this Financial Times piece, which inexplicably suggests the idea has some merit:

Mr McDonnell and Jeremy Corbyn, the new Labour leader, advocate a second approach: targeting QE at infrastructure projects. The central bank would buy bonds direct from the Treasury on the understanding that the funds would be used to improve housing and transport infrastructure. The timing is flawed; the Bank of England deems further QE unnecessary, and any large money creation now would risk stoking inflation. But if the idea were kept as something to implement the next time the country faces a financial crisis, it would carry quite a lot of respectability.

Some object that creating money to spend on infrastructure would undermine the central bank's independence by forcing it to buy direct from the Treasury. Yet monetary policy has already extended well beyond its technocratic bounds into the realms of wealth distribution. QE had clear wealth effects, which could have been offset by fiscal measures. All political parties should acknowledge this. So should those of us who want free markets to retain their legitimacy.


This idea is reminiscent of the Real Bills Doctrine, popular in the early 20th century. This doctrine tried to distinguish between central bank monetary injections used for productive purposes, and those that went into "speculation". Today economists consider this distinction to be meaningless, indeed there is no way for a central bank to control where its newly created money ends up.

Because money is fungible, it makes no difference if the central bank buys bonds that are used to fund infrastructure projects, or bonds that are used for some other purpose, say to fund military spending, or Social Security. All the money raised by a government ends up in the same pot, and it's impossible to say whether a specific dollar you received in a Social Security check came from funds the government raised through taxes, or buy borrowing from the Chinese, or by borrowing from the Fed. Nor does it make any difference whether the bonds are bought directly from the Treasury, or in the secondary market, where transactions costs are infinitesimal.

Conceivably the plan could make a difference if it actually caused the central government to produce more infrastructure. But that's a decision of the elected government, not the (unelected) central bank. The author notes (correctly) that the plan would lead to excessive inflation if adopted during a period where interest rates are above zero (as is likely to be the case when and if Corbyn actually took power.) Not mentioned is that QE could be combined with interest paid on reserves. But in that case the reserves would simply be another form of government borrowing, and hence the central bank would not have financed any government spending.

According to The Economist, Corbyn has already shown himself to be extraordinarily unprepared to lead a shadow cabinet. Looking at this proposal it's easy to see why. Even the Labour MPs seem embarrassed by his incompetence during "question time." They predict Corbyn will be replaced before the next election, to prevent a debacle for the once proud Labour Party.

PS. The FT article also claims that QE helped the rich. So what? Any policy that speeds a recovery from recession will help the rich, and the middle class, and the poor.

HT: Stephen Kirchner


Comments and Sharing






COMMENTS (17 to date)
Hazel Meade writes:

How is that a "People's QE"? If you want to really make it a People's QE you would simply write everyone cheques. This would be way more efficient and fairer. Everyone can spend the money on whatever they want.

Curtis Weems writes:

The suggestion seems similar to the Chinese government methods. Newly created People's Bank of China money enters primarily through manufacturing markets instead of exclusively through credit markets.

Mike sproul writes:

There is a good reason why the real bills doctrine says that money should be issued in exchange for real bills. It assures that money will be issued to carpenters and farmers, as opposed to gamblers and tourists. That way, the money supply will be elastic--rising when carpenters and farmers are busy, and falling when Business is slow.

Banks that issued money in exchange for fictitious bills (the kind offered by gamblers and tourists) will not only fail to provide an elastic currency, but they will also get caught in speculative bubbles and risk insolvency.

Remember: the real bills doctrine was developed by practical bankers over centuries of experience. It survived because it worked, and it is a better rule for monetary policy than any of the ivory tower theories that have tried to replace it.

BZ writes:

I simply don't understand how this can be true. Doesn't the Law of Supply directly imply real effects of where new money goes? First receivers may not matter in the long run, but for the duration of the money injections, how can they NOT shape the economy?

e.g. If the fed injected newly created funds, say billions per year, and spent it ONLY on lima beans, would not more farm land be dedicated to lima beans until the injections stopped? And isn't that re-diverted crop land a REAL effect that also affects Tractor sales, immigration, and lots of other things?

I am honestly perplexed by this post.

MikeP writes:

...a period where interest rates are above zero (as is likely to be the case when and if Corbyn actually took power.)

Why? Is there some correlation between interest rates and pigs flying that I am unaware of?

ThomasH writes:

The devil is in the details.

If the "bank" were just a way around the penchant of politicians to cut public investment during recessions, it would make a lot of sense. And if the monetary authority is politically constrained in the amount of LT bonds it can buy (and the "bank's" debt do not fall under this constraint), the additional debt would help even more. Ultimately the "bank" cannot do more than a government following a NPV rule for expenditures with present costs and future benefits, but few governments have followed that rule during the present period of below-trend NGDP.

Scott Sumner writes:

Mike, The problem is that the real bills doctrine leads to a procyclical monetary policy, making the business cycle more unstable. The monetary base fell between October 1929 and October 1930, which in retrospect was a disaster. The Fed pointed to less "need" for credit in a weak economy.

Having said that, yes, there are times you want to accommodate an increased demand for money. But that's not a general rule, just something that might be useful on occasion.

BZ, The mistake you are making is assuming that the monetary injection causes infrastructure spending to be higher. But it doesn't. That spending is determined by the legislature. It doesn't matter whether the central bank's new money buys those infrastructure bonds, or buys other bonds, and hence frees up other central government money for infrastructure.

Mike, Cute.

Thomas, I don't really understand your point. Yes, you could design some crazy system where it mattered, but why in the world would you make that assumption? In Britain the Prime Minister (or the Chancellor?) gives the BoE its mandate. This won't cause there to be more infrastructure.

Mike Sproul writes:

Scott:
The real bills doctrine leads banks to issue money when it is wanted, and to soak it up when it is not. Rather than exaggerating booms and busts, the RBD just accommodates the natural ups and downs of the economy. Think of it as just allowing the invisible hand to work for money as it works with goods.

Suppose the economy is slowing because of bad weather. Unwanted cash will be piling up in drawers and mattresses, and people will return that money to the issuing bank in exchange for the bank's assets (some of which are real bills). The banks are merely soaking up unwanted cash, which is not recessionary.

Now suppose good weather causes a boom. People will need more cash to conduct their business, and people will start bringing real bills into the banks to exchange for cash. The RBD will guide the banks to issue the needed cash, and the boom can continue without being pinched off by some misguided tight money policy (like in 1929-30).

Lorenzo from Oz writes:

Lawrence White recently set out key points about the Real Bills Doctrine, which seems to have some tendency to mean different things to different folks. How (not) unusual in economic theory debates

Jon writes:

Scott, I agree with your reply to Mike. The real bills doctrine was about matching the supply and demand for money... And it ignores hoarding. That's a problem, but I think it fails a better thought experiment...

Consider using open market purchases of tbills as the policy instrument. Treasury bills are real bills. The treasury issues them to manage steady government expenses against irregular tax recipients. So if the fed is following the natural rate--has neutral policy, buying tbills is neutral. Okay, suppose the fed buys all of the tbills as the real bills doctrine allows. Is the policy stable?

We know it is not. Ergo, the real bills doctrine is false. It's false because as yields drop there is a substitution effect by the broader market into other debt instruments.

This is the same reason operation twist schemes fail...

Scott Sumner writes:

Mike, As Lorenzo points out there are multiple versions of the real bills doctrine, but I don't buy into any of them. Let me take the case of the gold standard, which is the one of relevance to the historcial debate.

Start with the fact that an economic boom due to real factors will tend to raise nominal and real interest rates, as Summers and Barsky showed a few decades back. The higher interest rates cause the demand for gold to fall, which makes the boom more inflationary than otherwise. That by itself is somewhat destabilizing to labor markets. Now add in that during booms banks (and/or the central bank) will tend to issue more banknotes, which act as a substitute for gold coins. That can make policy even more procyclical.

Jon, Good points.

Mike sproul writes:

Scott.
No, the RBD is about matching the amount of money to the issuer's assets. if a bank issues $100 and gets $100 worth of bonds in exchange, then the new assets exactly cover the new liability and money holds its value.

Your thought experiment of imagining the fed bought everything ignores the law of reflux. If new dollars are not needed in the circulation, they will reflux to the fed as fast as they are issued, and the supposed increase in the money supply never gets off the ground.

Mike sproul writes:

Sorry, my comment at 1:05 was meant for Jon, not scott.

Mike sproul writes:

Scott:
That chain of events you describe is pretty wobbly. Boom causes interest to rise, which causes inflation, which causes more notes to be issued, which causes more inflation. This time I'm the one who doesn't buy it.

If you're worried about gold, then let the bank peg the dollar to a cpi basket. $1=1 basket. A boom happens and people want more cash. They bring things of value (including real bills) to the bank, and the bank prints up more dollars. While getting new assetS in exchange. The needed dollars are issued so the boom can continue (booms are not bad things), and you still have $1=1 basket.

Nathan W writes:

Theoretically, it seems like this must obviously be true.

But when you buy shares of a financial company, don't the benefits accrue directly to the people who enjoy higher prices of remaining shares? Money is fungible, and so there should be some trickling (both down and up?) of the benefits. But as obvious as the theoretical truth that it shouldn't matter where the money goes, when reasoning from perfect markets, is the conclusion from what I consider here to be "common sense" is rather different: the effect should be greatest where it is first applied.

The question of whether "people's QE" for social housing, etc. would be any different than a government promise to spend the same amount of money on the same project, seems quite fair. Here, since you refer to the very same decision maker across all government spending (the government), it is very different from reasoning about QE which sucks up private investments (such as by buying a bank that could have gone under). When allocated to private markets, central banks decisions in where to apply the QE might have serious allocational effects.

Yes, money is fungible, but consider a more specific sort of hypothetical situation: the QE money ends up directed to one of two banks. The first bank has specialist knowledge of lending in the construction industry, whereas the second bank has specialist knowledge for lending in mergers and acquisitions. Despite the theoretical equivalence in "perfect markets", I think "common sense" tells us that directing more QE to the bank with specialist knowledge in construction lending will have very different impacts on the economy than QE funds which effectively prop up the bank which specializes in M&A. Why? Because no entity is a perfect substitute for another, and even if the non-receiving bank factually could make "better" loans, the receiving bank will not be aware of these (will not receive the market signal, or at least, would receive only a heavily muted market signal).

Scott Sumner writes:

Mike, Sure, if you peg the CPI then there is no problem. My point is that if you don't have any regime in place to stabilize prices (and a gold standard obviously doesn't stabilize prices), then the RBD will make the business cycle more unstable, by making monetary policy more procyclical. The money supply will tend to rise during booms, and velocity will also rise, making NGDP more unstable.

Nathan, You are confusing money fungibility and the so-called Cantillon effects. If someone gives you $20,000 dollars then it makes absolutely no difference whether that action leads you to spend that exact $20,000 on a new car, or a different $20,000 that you already had in your bank account on a new car. That's what the fungibility of money refers to.

Mike Sproul writes:

Scott:
The thing that stabilizes prices is for the money-issuing bank to hold enough assets to cover the money it has issued. If the bank has issued $100, and holds assets worth 100 oz of gold, then the dollar will be stabilized at 1 oz. (Gold, as you noted, can still change in value.) If the bank holds assets worth 100 CPI baskets, then the dollar will be stabilized at 1 basket.

The RBD requires the bank to issue money in exchange for short term real bills of adequate value. "Short term" helps the bank avoid maturity mismatching. "Real bills" helps the bank provide an elastic currency without getting caught in speculative bubbles. The bank will issue more money when carpenters and farmers need it, but not when people want it to buy tulip bulbs and shares of South Sea stock.

"Adequate value" is the most important of the RBD's requirements. (and the most ignored by RBD critics) It assures that if the bank issues another $20, the bank will get assets worth 20 CPI baskets, and the price level will be stable. This does not destabilize the economy. It merely accommodates the economy's natural ups and downs.

Comments for this entry have been closed
Return to top