Sunday, December 9, 2012

Selgin on Cantillon Effects

George Selgin decided to defend Sheldon Richman's version of Cantillon effects against Scott Sumner.   I think he does a rather bad job of it.   Like many self-described Austrians, (which Selgin is not,) he seems attached to what I consider a confused and confusing framing of the matter.
The specific claim to which Scott objects is Sheldon's assertion that Fed open-market operations benefit those directly involved in them more than others because “early recipients—banks, securities dealers, government contractors—have the benefit of increased purchasing power before prices rise.” On its face that assertion appears, to me at least, as incontestable as the claim that, if I choose to do my Christmas shopping at Macy's rather than at Bloomingdale's, Macy's gains more from my display of Christmas spirit than Bloomingdale's does.

Nor can I see why it should matter whether the spending in question involves newly created money. Were I a counterfeiter whose products were perfectly indistinguishable from the real things, Macy's would still profit more from my spending than Bloomie's, provided it succeeds in fobbing the notes off just as easily as I do. The suppliers that Macy deals with are likely to profit as well, as may others who experience an increased nominal demand for their goods at stages of the "circular flow" not far removed from the fake notes' point-of-entry. Eventually, though, the notes will have worked their influence on prices generally, so that increased revenues, rather than going hand-in-hand with enhanced profits, merely compensate their recipients for a heightened cost either of living or of doing business.

I don't doubt that high-end retailers benefit by having more customers.   That is an aspect of imperfect competition.   The typical firm has excess capacity and profits from increased demand.    If one customer goes to Bloomingdale's rather than Macy's, then Bloomingdale has the added profits and Macy's does not.

However, this effect has little do with whether Bloomingdale's receives the money before prices have risen.   Suppose the quantity of money is $1000 and it increases by $100.   This is announced and all prices immediately rise by 10%.   The Fed then spends the money at Bloomingdale's rather than Macy's.   Bloomingdale's benefits almost exactly much as it would have before.   This is equivalent to an $90 increase in sales before the inflation.   Sure, if we imagine that somehow the stockholders of Bloomingdale's manage to get the profit and spend them on consumer goods before prices increased, they would gain a bit more--approximately 10% more.   But they receive 90% of the gain even if they only spend the profit from their mark-up after all prices have risen.

If the quantity of money had increased by $10, and prices immediately increased by 1%, then Bloomingdale's would benefit by the increased sales--approximately 99% of the mark up that would have occurred if the stockholders could have purchased the consumer goods before prices rose.   

No doubt primary security dealers are less than perfectly competitive.  But it is not at all clear that any benefit primary security dealers receive from trading with the Fed has anything do with getting the new money first before prices increase.   They buy and sell securities, including selling them to the Fed.   If prices rose immediately when the quantity of money increased, so that when the stockholders or partners get ready to spend their profit on consumer goods, they already have to pay higher prices, they still gain nearly all of the profit.

Further, consider an open market sale.   The primary securities dealers buy securities from the Fed which they sell to other investors.   Do the primary securities dealers lose money because they give up the money first before prices fall?  Or do they still make money on the transactions with the impact on the price level being trivial?

Or consider a scenario where the Fed churns, buying and selling bonds every day, leaving its net holdings and the quantity of money unchanged.   (Not an unrealistic scenario, sadly.)   Do the primary security dealers benefit from this?     It seems obvious to me that whatever profits primary securities dealers earn from trading with the Fed has approximately no connection with getting money first before prices rise.

It seems to me that emphasizing the notion that those receiving the money first spend it before prices rise is confusing.   It exaggerates a trivial aspect of the impact of the differential impact on demands.

Further, when the Fed calls up one primary security dealer to buy some bonds, then other clients are going to suffer some delay in getting their business done with that dealer.   Some of those clients switch to other primary security dealers who aren't  currently too busy with the Fed.

If the primary securities dealers were perfectly competitive, then this effect would be so important that it wouldn't matter which securities dealer gets the business of the Fed.   The other securities dealers would get enough other clients that it would make no difference.    All of the securities dealers might benefit from an extra demand for their services, but with perfect competition, there is no differential benefit to dealing with any one client.

This implies that the particular primary security dealer that gets the "new money" from the Fed does not benefit more than the other primary security dealers that get "old money" from investors turned away from the securities dealers busy with the Fed.

Suppose Bloomingdale's is already packed, and the Fed joins the line to get in.    Other customers, seeing the long line, go over to Macy's.     Does Bloomingdale's really do better than Macy's?    It might depend on whether the other customer that was deterred from Bloomingdale's was planning on buying some high mark-up item.     Of course, if both stores are nearly empty, then this isn't an issue.

More importantly, while Bloomingdale's extra profit involves "new money," paid by the Fed, the customers that switch over to Macy's are using "old money," and so, Macy's is earning added profit from "old money."    Focusing on the "new money" is a mistake.

Rather than think about two department stores, suppose the Chinese government announces a plan to purchase  5 million additional tons of wheat from U.S. farmers.    This raises the demand for wheat and results in higher wheat prices.   That the Chinese government is  buying wheat rather than corn makes a big difference to the wheat farmers.   But the particular farmers who sell wheat to the Chinese don't benefit more than those farmers who sell their wheat to the Japanese, the Europeans, or other Americans.   The demand for wheat rises, the market price of wheat rises, and all of the farmers benefit the same, regardless of to whom they sell their wheat.

Further, it is unclear whether the wheat dealers benefit much at all, much less that the ones who actually handle the Chinese sales benefit more than the ones who continue to buy wheat from American farmers and sell to the Japanese.  

What is the difference?   Wheat is more homogeneous than the services provided by retailers.   Further, we can imagine Bloomingdale's or Macy's selling off excess stock after Christmas at a loss.   It happens all the time.   Farmers sell all they want at the market price, rather than quote a price for their particular wheat and wait and see how many buyers show up to their farm.   If sales turn out a bit slow, they don't generally put on an after Christmas wheat sale.

When the Fed purchases T-bills, what happens?    Does it go to one primary dealer and buy all it wants from that dealer rather than another?   Is it like going to Bloomingdale's rather than Macy's?   Or is it more like the Chinese government announcing a purchase of wheat?   Do the T-bill prices rise, so that those selling T-bills to the Fed get the same price as those selling T-bills to other buyers.  In other words, is it similar to asking whether the farmers selling to the Chinese get a better price than the farmers selling to other buyers?

If it is more like an increase in the demand for wheat, then those selling T-bills to other investors for "old money" get exactly the same benefit as those selling to the Fed for "new money."   Tracing the new money through the economy  and counting every receipt of "new money" as increased demand and added profit, is simply a mistake.

What about the primary security dealers?   Do the particular primary security dealers buying from investors and selling to the Fed benefit more than the primary security dealers buying from other investors and selling to other investors?    To say that the primary securities must benefit from selling to the Fed or else they would not sell in no way shows that those securities dealers selling to the Fed benefit more than those who sell to other investors.     They must be benefiting from selling to other investors too, or they would not do it.

Finally, the U.S. runs budget deficits and so the Fed auctions off Treasury bills, notes, and bonds to the primary security dealers all the time.   The primary security dealers then sell those Treasury bills, notes, and bonds to investors.    Does it really make much difference to the profits of the primary security dealers if they sell the T-bills to the Federal Reserve rather than to some other investor?   Most importantly, do the particular primary security dealers that sell the bonds to the Federal Reserve open market trading desk make more profit than the primary security dealers who sell the bonds to other investors?

Focusing on the particular primary securities dealers that trade with the Fed because they receive "new money" that they just turn over to the Treasury seems like a mistake.    Assuming the Fed instead purchases T-bills from some private investor, focusing on that seller of T-bills because they receive the "new money" rather than all of those other people who also sold T-bills for "old money" or even held on to them, is a mistake.  And, identifying any benefits from the change in the pattern of demand with a minimal effect because money is received before prices rise is a mistake.

These mistakes--tracing the new money specifically and focusing on whether buying prices have risen, creates confusion.   In my view, the relevant effects, the element of truth in the "Austrian" story, is that excess supplies of money can differentially impact the pattern of demands.    If this happens, it will typically impact relative prices (leaving aside perfectly elastic supplies of goods) and almost certainly impact the composition of output (leaving aside perfectly inelastic supplies of goods.)     If these differential impacts on the composition of demand are expected to persist, there is an incentive to commission specific capital goods based upon the pattern of demands.   However, when the price level rises enough so that there is no longer an excess supply of money, then there is no more differential impact on the pattern of demands from the now nonexistent excess supply of money.

Worse, the focus on the pathway of the new money confounds a fundamental truth--an issuer of money obtains something for nothing--with the disequilibrium impact of an excess supply of money.    Even if prices instantly adjust so that the real quantity of money equals the demand for money, the seignorage  tax creates a revenue that can be spent.   Perhaps it is collected privately--by gold miners or counterfeiters.   Such a revenue or income can be spent and have a differential impact on demands.   As usual, changes in demands can change rents or quasi-rents.   These are secondary impacts of the seignorage.   This tax and revenue can be persistent.   It is not about patterns of spending that occur because prices haven't risen to their final equilibrium levels.

In my view, "Austrian" economists need to be careful to avoid these confusions, and further, to avoid communicating them to libertarian intellectuals, like Sheldon Richman, or ordinary citizens reading blogs or magazines.    The claim that those receiving the new money first benefit by spending it before prices have risen leads to confusion.


  1. Very nice post! I cannot agree more.

    2 points I can highlight:

    In your examples, there is one point which you failed to notice, although I perfectly agree that all holders of T-bills would benefit from rising prices, if competition is not perfect, those who buy to the fed earn a bit more for the simple reason that they get more volume (This could be debatable, although with the amount of T-bill the fed is buying this is probably true). But surely this impact is tiny when speaking of QE, the commissions earned by primary dealers are surely not having a macroeconomic impact of the order of magnitude that QE is doing.

    Second, if the T-bill market is actually a monopoly, or a monopolistic competition, then perhaps the fact that the fed goes to one dealer and not the other is benefitting this dealer more, but the T-bill market is one of the most competitive in the economy, market forces are quite strong in that domain, and surely one dealer cannot be considered a monopoly.

    I think where the "Austrians" have got it wrong, is that they think that QE is only having an impact through those small increase of profits, rather than the impact of inflation, which is coming more from an expectation perspective.

    So anyway to conclude, if the fed wants to purchase assets, it makes more sense to purchase assets in the least monopolistic areas - which in my view the T-bill market fits pefectly this definition.
    So Fed keep away from buying Macy's assets!

  2. If there is slack in the economy, then an increase in money should lead to an increase in output, not prices.

    What am I missing?