Wednesday, February 27, 2013

Sumner on Monetary Disequilibrium

Scott Sumner doesn't like the concept of monetary disequilibrium.  

Since Sumner's prefered transmission process for changes in the quantity of base money is "the hot potato effect," I think he cannot help but accept the monetary disequilibrium approach.  

Sumner claims that disequilibrium is a "psychological concept."    Apparently, the "psychology" must be that some individual feels discomfort.   Of course, the entire monetary disequilibrium approach is based upon the distinction between an individual who can easily adjust the actual quantity of money held to the amount desired, and the economy as a whole, which must somehow adjust the desired amount to be held to the existing aggregate quantity of money.    

The second step of the process is that when one individual spends more to be rid of excess money balances, someone else ends up with the excess balances.   This is "the hot potato effect."   If the focus is solely on the one individual who has already adjusted actual balances to desired balances, then that person has no pschological dissastisfaction.    But whoever now has the added, and now excess, balances has the disequilibrium.  (With excess money balances, this is a happy condition, a bit like a seller in a market with a shortage.   Plenty of customers--should I raise prices, produce more, or both?   Is it a problem?  Or an opportunity?)

But when he argues that there is no monetary disequilibrium, "the hot potato effect" is no where to be seen.   He traces though a plausible story by which an economy assumed to be in equilibrium responds through time to a "bolt from the blue," permanent change in the quantity of base money.  

Sumner describes the immediate effect of the permanent increase in the quantity of base money by what appears to be the "liquidity effect."   The traditional Keynesian approach would be that people spend any excess money balances on assets.   This causes asset prices to rise and asset yields to fall.    Base money is assumed to bear no interest.   The lower yields on other assets reduces the opportunity cost of holding base  money.   The amount of base money people want to hold rises.     Once asset prices are high enough, and asset yields are low enough, the quantity of money that people want to hold in aggregate matches the new, increased quantity that exists.

Given that approach, it is common to assert that there is no monetary disequilibrium once "the" interest rate has adjusted.   However, "the" interest rate is now below the natural interest rate, which leads to an amount of investment greater than the amount of saving.  (This could occur through more investment, less saving, or a bit of both.)  An alternative framing is that the sum of consumption and investment spending  is greater than potential output.    Shortages of output lead to higher prices.   This increases the demand for money, which results in higher market interest rates.   Once the market rate rises back to the natural interest rate, saving equals investment and the sum of investment and consumption matches the productive capacity of the economy.   The increase in base money resulted in a proportional increase in both the price level and nominal income.

Sumner usually dissents from this neo-Wicksellian/Keynesian approach.    I think a key problem is that rather than a liquidity effect where people with excess money balances just bid up asset prices and so reduce asset yields until they are willing to hold the existing quantity of money, Sumner is focusing on equities.   The long run effect of the permanent increase in the quantity of base money on the price level is that for any real level of profit, nominal profits will be higher.   This makes the fundamental nominal value of stocks higher.

Stock markets have very flexible prices, and so the market prices of the stocks immediately adjust to their new, higher fundamental values.   Since other goods and services have sticky prices, the real price of stocks  rises, but this is only until other prices catch up.    The nominal stock prices don't fall, but when other prices, such as the prices of consumer goods and services rise, real stock prices fall back to their initial value.

It is true that an exogenous and permanent increase in base money should immediately raise nominal stock prices.   And until other prices catch up, those already holding the stocks are "really" richer.     However, what does this have to do with monetary disequilibrium?   Yes, those who were owning stocks are temporarily richer, but does this cause a temporary increase in the demand to hold money?    

Is the assumption some fixed ratio between desired real money balances and actual real stock holdings?  

Of course, Sumner's usual approach is that "the hot potato effect" results in more expenditures on output directly and on labor indirectly.  This suggests that the temporary increase in real equity prices and the consequent greater real wealth does not increase money demand enough to clear up the excess supply of money.   Suppose nominal stock prices adjust to their new, long run equilibrium values and there remains an excess supply of money.   It would seem like anyone paying more than that price for the stocks would be setting themselves up for a capital loss.    People may have excess money balances, and they could buy stocks and bid their prices up more, but they don't want to buy stocks because their prices would be "too high."   So what do they do with those money balances now?

No, we are simply left with the simplistic focus on the individual that has already gotten rid of the hot potato being in equilibrium, while ignoring that whoever received the excess money now faces the disequilibrium.  (My own view is that the liquidity effect is real when the interest rates paid on money balances are fixed or sticky, but interest rates being different from the natural interest rate is monetary disequilibrium.   Saving greater than investment or total expenditure greater than potential output is monetary disequilibrium.)

Sumner then goes on to argue that the disequilibrium is in labor markets.   There is no imbalance between the quantity of money and demand to hold it, rather there is an imbalance between the quantity of labor supplied and demanded.   With a permanent increase in the quantity of base money, the disequilibrium would be firms feeling discomfort due to their inability to recruit workers.   (Actually, Sumner seems to believe that forced overtime is the problem.)

Is the problem "really" a shortage of labor?   Would an increase in the payroll tax paid by employers solve the problem?   In my view, the "solution" is for prices and wages to rise enough fso the real quantity of money falls enough to match the amount of real balances people want to hold.  To me, given that this is the solution, the problem is an excess supply of money.   Disequilibrium is an imbalance between supply and demand.

Also, this thought experiment of a permanent increase in the quantity of base money as a bolt from the blue is a special case.   Suppose the increase in the quantity of base money is not permanent and there is no permanent increase in future nominal income.  This entire step of stock prices rise immediately because of the long run impact on nominal future profit disappears.    But can't a temporary increase in base money cause a temporary excess supply of money?

Further, Sumner is misled by his emphasis on hand to hand currency.   He writes:

 When the Fed increases or decreases the base I do not walk around weeks later frustrated that my wallet contains $260 in cash, rather than $240 or $280. I’m holding exactly as much cash as I prefer to hold, given current asset prices.

Note the assumption that the Fed increases base money by handing out wads of currency to individuals.   Do they feel frustrated by carrying around excess currency?   Does Sumner personally have the experience of carrying about more currency than he wants?

Well, in reality, there is an easy way to get rid of excess currency.   Deposit it in a checkable deposit account.  So, sure enough, most individuals do not feel that they have too much currency.    Of course, most people would spend currency rather than use a debit or credit card or write a check.   But that shifts the currency into the hands of retailers.  And what do the retailers do?   They deposit it in their banks.   

Now, of course, the excess supply of base money is the hands of banks.   Is the problem that banks feel that their vaults are too full?   Well, banks constantly send worn currency to the Fed and order new currency.   All they have to do is order less new currency.   But then, the excess money balances are in the form of bank reserve balances a the Fed.

In reality, increases in base money directly increase the checkable deposits of people selling bonds to the Fed and the reserve balances of their banks.   And so, thinking about people walking about with wads of currency in their wallets is misleading.

To take Sumner's remark literally, however, he seems to be saying that people with excess currency in their wallet go to their stock broker and use it to purchase stocks or bonds.   And once stock and bond prices have risen, then they are happy to keep more currency in their wallet.  Really?   I am pretty sure that my average wallet holdings are not positively related to the S and P 500.

As for the money balances that actually count (the money in my checkable deposit,) I cannot begin to pretend that I am always holding the optimal amount.   Of course, I never feel "frustrated" as in, "unhappy" when my checking account balance is "too high."   No, I am pleased.   I happily solve this "problem," by spending more.   And, by the way, my money holdings are not strictly proportional to the S and P 500.

I have never had "forced overtime," so I don't really know how that would impact my desired money balances.    What I do believe is  aggregate desired nominal money balances--checkable deposits, currency, and bank reserves, is positively related to aggregate nominal income.

Perhaps it would be better to go back to thinking about a shortage of money, rather than a surplus.   But trying to identify "monetary disequilibrium" with people walking about with overfull wallets and having nothing to buy is wrongheaded.


  1. The "hot potato" effects is named because that's what it would like to an economist tracking the velocity of money after an increase in the money supply.

    It is not what it would feel like to possessors and spenders of the new money. They just feel richer , or feel that they now have more cash than they need to hold, so they spend a bit more.

    It seems valid to call this "monetary disequilibrium" because until prices adjust monetary spending in the economy will be above its eventual equilibrium level because ms > md.

    This additional spending will almost immediately drive up some prices but because of some prices being more flexible than others there will be some unevenness in the economic effects during the adjustment period. However precisely what these effects will be seems indeterminate to me.

    It certainly seems possible that Scott is correct and that we would would almost immediately move back to a position where ms= md, but where some prices (the most flexible ones including bonds and stocks) have gone up more than there eventual equilibrium while others (the more sticky ones including labor) have gone up less. But as this would indicate some irrationality on the part of equity buyers this seems unlikely.

    It seems more probable that the "hot potatato" effect would last some time (even years) as the final effects get worked out.

    That could well be what we are seeing now. md > ms because 4 years after a sharp increase in md prices have not adjusted.

    In any case the correct response of the monetary authorities would be to adjust ms to md and not wait for prices to adjust.

  2. Good post, Bill. I agree with you.