Friday, November 9, 2012

My Daily Target Practice

Bob Murphy sent over an email and wrote:

Anyway here's target practice for you if you're bored:
The link goes to a commentary from Scott Sumner, where he seemingly is attacking Austrian business cycle theory, without mentioning it by name. He attempts to rebut the notion that it is important where central bank money printing enters a system.

A point is being made by the Richmond Fed and goes along with Austrian thinking as to the fact that money created by the central bank enters the capital goods sector first:
When the Fed injects money into the economy, the effects are not spread evenly. The first point of impact is the banking system, where the Fed trades newly created money for assets. The infusion of cash causes financial institutions to bid down lending rates, which pushes down other lending rates in the economy and, the Fed hopes, stimulates the economy as a whole.
Sumner responds:
This is technically correct, but it creates a misleading impression.  I notice that many commenters believe it matters where the money is injected.  Not true.  If the Fed injected the money into the computer software industry by buying T-bonds from Microsoft, the impact would be essentially identical.  Microsoft would probably take the cash and deposit it in the bank the same day. But even if they didn’t, even if they spent the money on stocks, the impact on interest rates would be identical.

What Sumner is missing here is that the Austrian view is that the money enters the capital goods sector distorting the capital-consumption structure in favor of the capital sector. Without realizing it, Sumner makes the same point. He is not refuting the key point about movement in favor of the capital structure.

Less accurate may be Sumner's further point:
Even before the Fed existed, easy money would often depress short terms rates. For instance, a sudden discovery of gold would depress interest rates under the gold standard. This is because monetary policy has nothing to do with credit, it’s all about changes in the stock of the medium of account (MOA.) When you increase the stock of MOA then prices should increase in proportion. But because prices are sticky in the short run, people temporarily hold excess cash balances (or excess gold balances under the gold standard.) Since the nominal interest rate on the MOA is zero, the nominal interest rate on financial investments is the opportunity cost of holding the MOA.
I'm not sure Sumner is historically accurate here. Can he actually name a period where a gold discovery pushed interest rates down? Can he show that in all historical instances this has occurred? I'm not saying it can't occur in some individual case, but I don't see any stream of logical reasoning that would prove this must occur under all circumstances when a major new gold discovery is made. The gold could just as well cause an increase in consumption spending, rather than it flowing into the capital goods sector.

Sumner than throws out that "prices are sticky in the short run." Huh, tell that to people currently trying to buy gas on the black market in NYC? But further, Sumner fails to make the important distinction between saving and holding a cash balance. Saving, in my view, is money that is used to buy capital goods or put out on the loan market. This could occur after a gold discovery and would push interest rates down. However, the possibility also may exist that the gold is just held as part of a cash balance, which would have no impact on interest rates or prices. This Sumner doesn't seem to get and suggests that a desire to increase cash balances will push interest rates lower, when they will either keep interest rates the same (when their is a new gold discovery) or push rates higher (when it removes funds from the capital sector). Thus, Sumner is failing to make the distinction here between saving and an increased desire to hold cash balances and the differing impacts on interest rates and the capital goods sector.

Then, of course, there is the very real initial point made by me that the cash could just spent on consumer goods, which Sumner seems to rule out from some unknown historical data perspective and also because "prices are sticky"--which seems to me not even, as viewed by Sumner, as prices really being sticky but as gold finders not desiring to spend their new found gold. Contrast this with what we see about those that win lotteries, and not spending seems to be the opposite of their problems.

Sumner then goes on to make declining interest rates the core causal factor behind an increase in the desire to hold cash balances, which adds nothing to his argument but further confusion, since he makes this case outside of gold having any role in this movement  In short, start to finish we have some very confused thinking by this guy Sumner.


  1. A more simple explanation of the fallacy in his argument is this:

    If the Fed bought bonds from Microsoft, MS would probably immediately deposit that money in the bank. However, the reason they were selling the bonds in the first place is because they were intending on using the money for some planned project. Therefore, before that deposited money has a chance to spread throughout the entire system via the money multiplier effect, Microsoft got to spend a lot of that money BEFORE prices throughout the economy.

    It does matter where money enters the system. Those who get it first can spend it first before the new prices stabilize.

  2. "Sumner then goes on to make declining interest rates the core causal factor behind an increase in the desire to hold cash balances.."

    Wouldn't declining interest rates factor behind an increase to spend cash balances? It pushes people to spend their money rather than lose it to inflation.