Bryan Caplan  

Hummel on Central Banking

Antitrust for Hong Kong?... Freddie Mac: The Book...

Hummel follows up on my post on the (relative) efficiency of modern central banks:

Here are my three reasons, given somewhat sporadically in my lecture, for the better performance of central banks in developed countries since the 1980s:

1. Highly developed financial systems with widespread fractional reserve banking have reduced government seigniorage, even at double-digit inflation rates, to a trivial source of revenue...

2. Globalization and international competition have approximated Hayek's world of competing private banks issuing fiat money...

3. Central banks are freer to respond sensibly to this growing international competition and market discipline because of their political independence.

All fine points. My main quibble with Jeff:
Economists only consider inflation's deadweight loss, ignoring inflation's transfer, which bothers the public just as much and just as reasonably as the transfer from the income tax. Does it really make sense to say that the public hates taxes too much because most of the extracted revenue is just a transfer?
Since most inflation is anticipated, I don't see that it does transfer much; instead, it's built into raises and interest rates. But many members of the public have a different - and crazy - model. They imagine that if inflation were 0%, their nominal income would continue to rise at its current rate.

Furthermore, when inflation is unanticipated, many people ignore all of the transfers they receive. If you have a fixed rate mortgage, double-digit inflation is probably great for you. But how many home-owners who lived through the inflation of the 70s even stopped to think about how much inflation had done for them?

To repeat, I think it's good that people overestimate the damage of inflation. It's one of the main things that makes modern central banks so well-behaved. But in purely factual terms, the public's beliefs about inflation are sorely mistaken.

P.S. Check out Jeff's piece on inflation aversion in Econ Journal Watch; it looks quite good.

Comments and Sharing

COMMENTS (15 to date)


What do you say about the greater deadweight losses that occur from bubbles, that go unanticipated by FED POlicy makers? The economy is so complex and inter-tied that its difficult for the FEd to know when they are about to screw up. The housing market is one. Bernanke prolly didn't know that increasing the the interest rate would screw with complex derivates backed by subprime mortgages.

dWj writes:

I remember reading a comment by someone -- I think it was Bill Poole, but I'm not certain -- that economists tend to underestimate the cost of (anticipated) inflation because they and the people they associate with are good with natural logarithms.

Libra writes:

Bryan - I think you're missing the point about the losses from inflation.

Anytime the money supply increases by a certain percentage, you dilute current money holders by that percentage. If there is $100 billion in the money supply, and I have $1,000, printing out $5 billion and giving it to Uncle Sam makes me 5% poorer, and Uncle Sam gets $5 billion for nothing.

In the past decade the money supply has been growing at least 5% a year (I'm using the figures for nominal GDP growth). If this was a neutral distribution, I could keep $100 in a checking account, and every year I would get $5 from a helicopter. However, that is not happening. I can't even earn a 5% return from a savings account or a CD. So tell me, how is that 5% getting created? Who gets the money? The money must enter the system in some way.

I believe the answer is that the money goes to everyone involved in the fractional reserve system. Some of it goes to Wall St. bankers, some of the money goes to debtors ( homeowners), a bit of the money goes to people with savings accounts, and tiny amount goes to seigniorage.

The point is there is a massive transfer going on. There is simply no way to grow the money supply continuously at 5% a year without transferring money to the creators. Just because the transfer isn't going to government does not mean it's not a problem. There is no justice in having people with checking accounts transfer their money to Wall St. and homeowners.

Second, there is a massive dead weight loss. The high rates of monetary growth have led people to invest far more than they should in property and stocks ( as opposed to keeping their money in dollars). The overproduction of internet stocks and McMansions is a case study in malinvestment.

Adam writes:

You're only right about inflation having no impact if it was instantaneous. But it's not, it takes several months or years for inflation to fully percolate through an economy. So while it's true that a homeowner will gain from an inflation transfer, that transfer will not happen at the same time as any negative transfers they experience. It could be to their benefit or detriment, depending on the timing.

The way this works in practice is that there are groups of people very close to the top of the inflation pyramid (the government being the very top) who will benefit much more than those farther down. The net effect of inflation (expected and unexpected) is typically a wealth transfer up toward the inflation source.

Matt writes:

Inflation Aversion and the poor

The problem with inflation for the poor are quantum problems, the inability to get efficient arrangements of lot sizes to fit changes in money value.

We can only buy eggs in units of a dozen, cutting them up to half dozen is very inefficient. Poor people in grocery stores have fewer lot size choices than rich people.

In grocery shopping alone, poor people pay a high price in incomplete lot sizes. At their end of the market, fewer choices.

Why not plan linearly in time for linear inflation? Impossible, Instability Theory, even a linear predictive inflation will be cyclic.

floccina writes:

One thing that might be a problem to the poor from accelerating inflation is that the poorest among us cannot get credit and are thus put at a relative disadvantage.

ionides writes:

Suppose you went to a store to buy a couch. "I want a 10 foot couch". "That will be ready in 6 months". "6 months? Make it a 12-foot couch".

That's what inflation does. It makes it difficult to compare magnitudes across time. It introduces a temporal dimension to a non-temporal quantity.

This makes it hard to plan. It makes it hard to visualize future prices. How much will it cost to send a kid to college in 15 years? How much savings are required? Inflation makes it impossible to answer these questions.

Mike Sproul writes:


The belief that printing money causes a wealth transfer is wrong. If the new money were printed by a counterfeiter who did not put his name on the money, did not recognize that money as his liability, and did not stand ready to use his assets to buy back that money, then one could say, with the Ron Paul crowd, that the new money was "created out of thin air" and caused a wealth transfer (to the counterfeiter and the first recipients of the money). But a few minutes' reflection should convince you that the fed is not like a counterfeiter. It puts its name on its money, recognizes it as its liability, and stands ready to use its assets to buy back the money it has issued. Every time the fed issues a new dollar, it gets a dollar's worth of assets. Thus the fed's assets move in step with its issue of money, and the issue of new money is not inflationary.

Adam writes:
It puts its name on its money, recognizes it as its liability, and stands ready to use its assets to buy back the money it has issued. Every time the fed issues a new dollar, it gets a dollar's worth of assets. Thus the fed's assets move in step with its issue of money, and the issue of new money is not inflationary.

Mike, I think you're misunderstanding something here. What are the Fed's assets they get in exchange for their dollars? T-bills.

T-bills aren't real assets in the way that gold or commodities are. They're liabilities on the government (taxpayers). They are essentially created out of thin air. Your theory of inflation holds water if the money issued replaces of some other form of money, thus removing an equal amount from the economy as it puts in. Money backed by T-bills doesn't remove anything, therefore it has an inflationary effect.

Mike Sproul writes:


"Mike, I think you're misunderstanding something here. What are the Fed's assets they get in exchange for their dollars? T-bills. "

No. I'm stating the real bills view of money, while you are taking the quantity theory view.

T-bills have value just like anything else. As long as new dollars are issued in exchange for something of equal value, it makes no difference if that 'something' is US T-bills, British bonds, gold, farmland, or lottery tickets. The dollar is just like any other financial security. If a new issue is matched by an increase in the assets of the issuer, the value of the security (or money) will be maintained. If the new issue outruns assets, then the security (or money) will lose value

Bill Woolsey writes:

The dollar is defined in terms of the Federal Reserve note. The "price" of Federal Reserve notes, then, is the inverse of the price level. Since open market operations clearly change the supply of Federal Reserve notes, to claim that this has no impact on the price level, one must explain how the demand to hold Federal Reserve notes always changes in step with supply. That the Fed changes its holdings of Treasury Bills does nothing to guarantee a changing in demand to hold Federal Reserve notes.

The notion that interest bearing treasury bills are perfect substitutes for zero-interest currency and reserves that can be used to make payments is simply absurd.

In other words, the notion that a decrease in the supply of Treasury bills outide the Fed must create a matching demand for Federal Reserve notes is absurd.

It is simply false that firms can selll unlimited quantities of their own liabilities at a fixed price as long as they use the proceeds to purchase assets.

For a bond instrument, the promised interest rate is set in conjunction with the planned quantity of issue. Even if those plans were correct, and additional quantities are issued, what happens is that the bonds will trade at a discount from par. Their price will fall. This is true even if the firm is purchasing assets with the funds raised by selling the bonds.

If the institution is issuing base money, however, the can be no "discount from par." Instead, the price level rises.

Anyway, the demand to hold zero interest bearing instruments (like Federal Reserve notes) is limited. They make a poor store of wealth relative to assets that generate income. Over issue results in a decrease in their value (like any other security issued in "excessive" quantiteis relative to the demand to hold them,) and so the result is inflation.

Since the Federal government owns the Fed, and the assets of the Fed are liabilities of the Federal government, and, these liabilites are promises to pay the liabilites of the Fed, there is no real obligation of the Federal Reserve or the Federal Government to pay any external asset to anyone.

They are creating money out of thin air and using the proceeds to finance government projects.

The only limit on the activity is the political one--voters who punish politicians because they don't like inflation.

Mike Sproul writes:

Bill Woolsey:

"They are creating money out of thin air and using the proceeds to finance government projects."

No. They are issuing money that is backed with bonds, which are in turn backed by taxes receivable. If the government issues more money than its assets can support, then of course the money supply will outrun the assets backing it and inflation will result. But if the government issues more money, and at the same time earmarks more bonds as backing for that money, there is no mecessary tendency toward inflation.


In 1685, the payroll was late in arriving to a french fort in quebec. Soldiers were suffering, as were the shopkeepers who depended on their business.The payroll officer decided to issue paper IOU's, written on playing cards, that each said "1 livre", with the promise that when the payroll ship arrived, the possessor of the IOU could redeem it for 1 livre in coin. The IOU's circulated successfully as money, ending the bad economic times. Assuming the payroll ship contained 30,000 livres in coins, the payroll officer could safely issue up to 30,000 paper livres without causing inflation.
Now suppose that a soldier comes to the payroll officer and offers a french government bond with a market value of 100 livres, if the payroll officer will give him 100 paper livres. Another soldier requests a loan of 200 paper livres, and promises his farm (worth 300 livres) as collateral. Or suppose the payroll officer, in order to conduct business efficiently, needs a new desk, which he pays for by printing up 300 new paper livres.

In each case, the issue of new money was backed by assets of adequate value. In none of the cases just mentioned would the value of paper livres fall below a silver livre, even though it would be fair to say that the french government was holding its own IOU's (and those of the soldiers).

Adam writes:


You are in error about the value of a t-bill. A t-bill doesn't gain any value until after is has been purchased. Until then it is merely a piece of paper or an entry in a computer. The government could print up a trillion billion dollar t-bills and it doesn't mean that the government would then own $1,000,000,000,000,000,000,000,000 in assets.

Likewise with federal reserve dollars. They have no value until after they have been used to purchase something. Until then they're also just pieces of paper or entries in computers.

You reasoning is circular. You claim that federal reserve dollars have value because they have been used to purchase t-bills and that those t-bills have value because they were purchased with federal reserve dollars. Two things of no value don't suddenly gain value because they have been exchanged for one another.

Like the law of conservation of energy, the value must come from somewhere. In the case of federal reserve dollars, the value that is imparted to them comes from inflation.

If the government sold those $1,000,000,000,000,000,000,000,000 worth of t-bills to the federal reserve that would certainly be inflationary. So then why would them selling $1,000 worth of t-bills not be? Where is the magical dividing line between these two numbers?

Mike Sproul writes:


If I pay $1000 for a share in the Brooklyn Bridge, that share was worth zero before I paid for it and zero afterwards. But a Treasury bond is worth $1000 before and after I buy it.

Adam writes:


But a Treasury bond is worth $1000 before and after I buy it.

So you must think that if the treasury printed up one trillion $1,000 t-bills then the government would be the proud owner of one quadrillion dollars in assets. Interesting. Sounds like a great way to purchase all of the other assets in the entire world.

The mistake you're making is in freely converting between potential and real value. It's the same as converting between potential and kinetic energy in physics. Potential energy doesn't exist, it's only a theoretical concept used for making calculations. Real value, like kinetic energy, can only be imparted from one object into another, it cannot come from potential value.

Comments for this entry have been closed
Return to top