Friday, April 25, 2014

Observational Equivalence

I often see this posted as a critique of government policies. However, I think it better serves as an illustration of observational equivalence. In this case, is the town experiencing a balance sheet recession or an excess money demand recession? Both views can be inferred from the empirical facts. Readers of this blog will not be surprised where I come down on this.
It is a slow day in the small town and streets are deserted. Times are tough and the economy is struggling.

A rich tourist visiting the area drives through town, stops at the motel, and lays a $100 bill on the desk saying he wants to inspect the rooms upstairs to pick one for the night.

As soon as he walks upstairs, the motel owner grabs the bill and runs next door to pay his debt to the butcher.

The butcher takes the $100 and runs down the street to retire his debt to the pig farmer.

The pig farmer takes the $100 and heads off to pay his bill to his supplier, the Farmer's Co-op.

The guy at the Farmer's Co-op takes the $100 and runs to pay his debt to the local prostitute, who has also been facing hard times and has had to offer her "services" on credit.

The hooker rushes to the hotel and pays off her room bill with the hotel owner.

The hotel proprietor then places the $100 back on the counter so the rich traveler will not suspect anything.

At that moment the traveler comes down the stairs, states that the rooms are not satisfactory, picks up the $100 bill and leaves town.

No one produced anything. No one earned anything... However, the whole town is now out of debt and now looks to the future with a lot more optimism. 
So what say you? Is this a balance sheet recession or an excess money demand recession?

Update: Nick Rowe builds upon this discussion.


  1. Whatever the answer , it's a bad analogy for the current U.S. economy.

    Here's how you fix it :

    Rich tourist visits town , looking for the motel owner , the butcher , the pig farmer , the Co-op supplier , and the prostitute - each of whom owes him $100 , plus interest and late fees.

    So what say you? Is this a balance sheet recession or an excess money demand recession?

    1. That's simple: the motel owner accommodates him, the butcher butchers him and give the corpse to the pig farmer. The Co-op guy and the prostitute go out to celebrate.
      I would say it is a daring solutions recession

  2. Why wouldn't the butcher and pig farmer borrow the recently "repaid" $100 to expand their meat distribution business? Maybe they think the investment opportunity is poor (i.e. expanding their balance sheet for little prospect for return)? I think it is very difficult to pick one over the other; it's observationally equivalent as you noted on another post.

  3. This comment has been removed by the author.

  4. I have another version of the small town. A historical novel of sorts.

    In the Old West, a handsome stranger comes to town. He heads for the local saloon-hotel, and places a crisp $20 on the counter. "Drinks, food and rooms, until that it used up."

    The keeper pockets the money and hires a man to fix his leaky roof. The roofer hires a crew for a day to improve the drainage ditches on his small farm, the crew blows the money on some girls. The girls go back to the saloon and buy drinks for the house until the bill is exhausted again. The $20 bill for a few more weeks circulates through town, resulting in more infartsructutra improvement, product and services being generated.

    Then one day a US Marshall shows up in town, goes to the saloon counter. "Say, have you seen a handsome stranger here?" the marshall asks.

    "We do not see many marshals here," answers the barkeep. "Why do you ask?"

    "He is a counterfeiter," says the marshall. "Passing fake $20s."

    The barkeep ponders the no longer fresh $20 in his till.

    "Ain't seen nobody like that," says the barkeep.


    Or, you can have a second ending:

    "Yea, I have the bogus $20 in my till," answers the barkeep.

    The saloon is soon packed.

    "We have been robbed!" cry the roofers, the farmers, the barkeep, the girls and and other townspeople.

    Everybody is soon after everyone else "for real money."


    Why is this historical fiction?

    There was a lot of counterfeiting before 1913, due to the multiplicity of banks and bill issuers. Thousands of banks were issuing money. This counterfeit money helped develop the West--and there is lesson in that.

  5. Money is a clearing house, where circles of offsetting obligations are cancelled out.

  6. David: I think this is an important post. I have some further thoughts on it:

  7. The tourist provided incremental “liquidity” in the form of a deposit with the hotel owner.

    The deposit is simply the (contingent) transfer of a bank balance.

    That provides impetus for the subsequent chain-deleveraging of balance sheets.

    Excess money demand was a temporary freeze on money velocity – a local liquidity trap.

    Then money velocity suddenly increased as a result of the first transaction.

    That set things in motion.

    From a flow of funds accounting perspective:

    The monetarist excess money demand view frames things implicitly at the ultimate source of funds – money that is or isn’t being applied from the source. If it isn't being applied, that sets up for excess money demand.

    The balance sheet recession view frames things implicitly at the ultimate use of funds – where money is or isn’t being applied as a use. It it isn't being applied, that sets up for a recession due to an inability to deleverage from various uses of funds.

    The monetarist excess money demand view seem inherently bottom up.

    The balance sheet recession view seems inherently top down.

    The balance sheet recession view and the money view are two different perspectives on the same thing. They are both accounting frameworks for viewing the same dynamic. And they are the same thing when the accounting is fully integrated. It's almost a tautology.

    (Accounting is a framework for the measurement of the condition of balance sheet stocks and flows.)

  8. The fictional story always works fine up til one of the people being paid with the $100 was a net creditor.

  9. This is spot on! The (converse) collary is: debt was piled up in the town in the first place, because of a lack of "real money". If there were more $100 bills floating around the town earlier, then the debts would have been discharged much earlier.

    What you are saying is that tight -- not loose -- monetary policy creates debt and credit! Most of the profession has this exactly backwards.

    In my work, I use tight money / loose money examples of an isolated mining town - where everybody is paid/pays on credit during the snow season, until the spring bank stagecoach comes in with a load of "real money" that discharges debts. Monetary policy goes from tight to loose. Debt builds up during tight money periods, debt goes down during loose monetary periods.

    Also, debt traditionally gets liquidated by currency. Total US public and private debt since 1918 divided by the currency portion of the monetary base is something to behold.

  10. David, nice article. I'm glad Nick Rowe highlighted it... I almost skipped over it except for that. Question: your parable illustrates a tight money problem, and it's resolution. Is it possible to extend this same basic story to illustrate a problem in the other direction (money which is too loose?), and it's solution. If so, what's the most insightful way to do that? Any thoughts?

    I was thinking the other day that if I was given control of a central bank (and I could ensure I wouldn't be run out of town -- or worse -- by an angry mob) that even I (with my limited knowledge) might be able to break anybody's economy in at least two spectacular and different ways: with too tight a monetary policy (sell all the CB's assets, stop ordering new currency or exchanging CB deposits for it, bump IOR up to 10%, and raise the reserve requirement to 500%), or too lose a policy (buy the equivalent of 10x the country's GDP in a month, (maybe pay cash?), and keep going at one annual GDP per month, while lowering IOR to -10%, and the reserve requirement and overnight rate to 0%.

    What do you think? Do you think I could break economies like that, with either of those two extreme policies?

    If just one other policy between those two on the tight to loose money spectrum is better than my two extreme ruinous policies, then this implies a maximum "goodness" of policy somewhere in between, right? Then it's just a matter of finding it. In other words we could "prove" an optimal policy exists along the loose to tight spectrum. We can't prove uniqueness, but we can prove existence.

    1. Tom, the question is tougher than it first seems. For example, my knee jerk response is yes, the Fed could push actual interest rate (i) below the natural interest rate (i*)--only if i* is greater than 0--and that is what I think they did in 2002-2004. This was too loose policy. However, it is not clear that doing this should necessarily lead to the huge bust like we had over the past five years. As I have argued before, maybe a mild recession was inevitably but not the Great recession. The Fed made/allowed the Great Recession to happen (this time by allowing i* to fall below i). It wasn't inevitable. So I am beginning to warm up to the idea that the consequences of being too loose vs.too tight are asymmetric. This lends itself to Milton Friedman's Plucking model.

    2. Thanks for your response David. Bizarrely enough, this is the second time today I'm reading about the "plucking model"... here's the 1st instance (which caused me to do a little research to see what was being "plucked"):

    3. The way Jason Smith is thinking in terms of the plucking model is that expectations, either too pessimistic, or too optimistic, can destroy information in the market, which is a bad thing. And actually he thinks that it may be very difficult to approach expectations with a "maximally ignorant" prior... you can beat that with an expectation which is very accurate, but the consequences of being wrong in your expectations, either direction, can be large.

    4. It's not necessarily difficult to use the least informative prior as your model -- humans just have a tendency to be wrong when trying to predict the future :)

    5. Jason, sorry, I botched your idea by by not proof reading my comment... I meant to write

      "And actually he thinks that it may be very difficult to beat expectations with a "maximally ignorant" prior... you can beat it with an expectation which is very accurate..."

  11. To say that recessions are caused by "an excess demand for money" comes to the same as Keynes's point about the "paradox of thrift": i.e. if saving rises, then all else equal, demand falls.

    And that's a point with which MMTers have no quarrel.

  12. Hi David,

    Slightly Offtopic.(left the same question at Nick Rowe's place as well)

    It seems like the need for money is a positive function of the number of participants in the wicksellian circle, keeping communication, laws and trust constant, since every step will want to maintain a cash balance. Doesn't this imply that Nominal Gross output level targeting might be better than NGDPLT, since it will more closely reflect the demand for money for transactions?

    1. Prakash, I have been thinking a lot about Gross Output lately. Though it is no substitute for what GDP measures--value added economic activity--it does better measure total transactions. The original equation of exchange was actually stated MV=PT where T is transactions. Typically, we use MV=PY where Y is real income or real GDP. I do think there is something here.

  13. The $100 is a device to move the information of who owes money to whom around the town; the town had been suffering from inefficient information flow that likely could have been cured with looser monetary policy (more money). However the monetary base = $100 is now large compared to NGDP = $500, so an additional $100 likely would not have as big of an effect.

    1. Jason, you should appreciate this famous article, "Money is memory"

    2. Thank you for that reference! That's a good one. These two ideas might be the same thing. Interestingly, I managed to take the information idea and and match it up (sorry for the pun) with matching theory ...