Saturday, March 30, 2013

Two Artless Queries about the Zero Bound

Paul Krugman's "The Price is Wrong" is about as succinct an explanation as you are likely to get of his position of whether/why to stick with a policy of loose money and unbalanced budgets.  The analysis pretty much makes sense to me although I'm not nearly skilled enough to judge the empirical evidence in any detail.  I do at  least acknowledge that Krugman seems to be winning on points, insofar as the threatened boogieman inflation persists in snoozing placidly under the bed rather than  lurching randomly round the boudoir  and upending  all the furniture.

But two unrelated questions continue to nag at me.  One, the "zero bound" discussion rests on the premise that the nominal rate cannot go below zero (yes?).   But it does go below zero (yes?) even in ordinary times in,, e.g., any checking account that imposes a fee for holding the deposit?  An objector might say that this isn't price-of-money, it's a service/convenience charge.  I'm not persuaded that there is a difference, particularly given that economists are (so far as I can tell) generally a little shaky on explaining just what we are paying for when we pay  price for money.*

Anyway, I find it hard/impossible to believe that all depositors/investors are deceived by the positive-nominal, negative-real fandango.  At least some of that money must be coming from investors who know perfectly well that they are getting a negative return but figure they might as well suck it up and soger on anyway.  

Which bring me to my second question--the topic is the idea  that a little inflation might be a good thing insofar as it drives down real wages.  Apparently this notion has been part of the argument from the beginning: I recall reading it in a letter from John Maynard Keynes, quoted, I believe, in a book by Bruce Bartlett.  As an empirical proposition this may very well be true (why, for comparison, do consumers persist in carrying credit card debt at astronomical rates when they often have other and better choices?).  Yet isn't it strange to construct an economic policy program that only works if the consumer does not understand what is good for her?

[Note: Revised, to take account of some second thoughts and a bit of Larry's commentary, infra.  But basically impenitent.]

*My Uncle Evert used to say he left his money in his copy of Dante's Inferno, so when he asked himself "now where in hell did I leave my money?" he would know where to start looking.  Cypriots may feel the need to resuscitate and refashion this joke.  BTW in a digital, does anybody but me find it intriguing that the Brits are freighting actual pound notes into Cyprus to solace the troops?   I suppose if they simply dropped them from helicopters they might solve the whole problem.  

5 comments:

Larry Hamelin said...

I'm just an undergraduate economics student, but let me see if I can help.

One, the "zero bound" discussion rests on the premise that the real rate cannot go below zero (yes?).

No. It is nominal interest rates that cannot go below zero. Hence Krugman advocates just enough inflation so that we can have zero nominal rates and negative real rates (nominal rates less inflation)

But it does go below zero (yes?) even in ordinary times in,, e.g., any checking account that imposes a service charge?

That is, as best I can tell, not an example of negative nominal rates. You're paying the bank a fee for a service (keeping your cash safe and clearing your checks), not borrowing money and agreeing to pay back less than the principal amount.

I tend to let my checking balance ride a bit high, recognizing that I am paying a price, but knowing that I'm buying a form of overdraft protection?

The price you're paying, though, is not fees, but rather the opportunity cost of investing the extra money.

[A] little inflation might be a good thing insofar as it drives down real wages . . . has been part of the argument from the beginning.

Which argument? I just mean, there are a lot of arguments floating around, but Krugman doesn't talk about lowering the real wages, which seems like a Bad Idea in a recessionary gap.

Yet isn't it strange to construct an economic policy program that only works if the consumer does not understand what is good for her?

What, you don't like capitalism? ;-)

Larry Hamelin said...

That should read (with a semicolon preceding) "the bank is not borrowing money and agreeing to pay back less than the principal amount."

HTH

Larry Hamelin said...

Another note about using inflation to drive down real wages...

There's an argument that nominal wages, i.e. the dollar amount of wage, resist going lower (downward nominal wage rigidity).

There are a lot of reasons why this might be so; my favorite (which few economists seem to discuss) is that in the modern credit economy, people have nominal debts that don't follow the price level: even if consumer prices follow wage cuts, they still lose out because they have mortgage payments, car payments, student loan payments, etc. they have to make, which don't follow the CPI.

If the economy is in an inflationary gap, there's an argument that price inflation, if not accompanied by wage increase, could lower real wages enough to restore macro equilibrium.

However, as mentioned earlier, we are in a recessionary gap, so we want to raise real wages.

Ken Houghton said...

Larry's got most of it. Just take it to the next step.

The rate you see (i; nominal) is made of the real rate (r) and the expected inflation (pi^e).

This is fine for any situation where r or pi^e are positive, or any situtation where one or the other is negative but the negative is less negative that the position. (Deflationary environment: pi^e = -1, but real growth remains around 2%, so i=1. Or inflation and negative growth: r= -1 and pi^e=2, so again i = 1.)

Sidebar that pi^e will not necessarily be pi; only expectations matter.

If real growth is more negative than inflation--the economy contracts at 3% while inflation is only 2%, the only way to "clear" the market is to have i = -3 + 2 = -1 (-0.01 in numeric form)

But i is bounded at 0. (Those fees are not paid in most economic models--and note that if you have $100K in the bank, the drag of having to have $5K to avoid fees is negligible. This is a flaw, but the models only approximate reality.)

So there are two ways to achieve a clearing rate of 0: either (a) slow the contraction to 2% or (b) raise inflation to 3%. (In reality, you can probably only approach those levels and the other will adjust, which sometimes gets called the "multiplier effect" and sometimes is a change of "expectations." You say scallion, I say green onion.)

The point of the ZLB is that unless you can realize i, there will be market inefficiencies--it will not "clear" for some transactions, and potential will be unused. So growth in the next period (r') will be lower than it should be, and it will take longer to reach a point where i should be >= 0.

Either you use fiscal policy--reduce contraction so that r increases--but it takes some time (bills passed, funds allocated, bids put out, etc.) and may last longer than you need it to and therefore create the mythical "crowding out" (government uses construction workers so that businesses cannot hire them).

The other choice is to raise inflation expectations (pi^e). How do you do that? Print money and distribute it. The first part is much easier, quicker, and easier to reverse.

Distribution is something that happens in economic theory because banks are intermediaries, and make their money by lending money. That some brain-dead idiots persist in believing this remains true--oh, sorry, where was I? Oh, yeah, so banks are always solvent and would never sit on that "excess" money (let alone deign to collect interest on it from the Fed to cover their trading losses and hide their insolvency) but rather would circulate it which will raise inflation and again allow the market to clear.

Since the market can only clear completely at i>=0, we call it a ZLB. What really happens is that some things don't get done--investments are not made, businesses are not started, expansions are delayed--that could if the contraction were reduced and/or inflation expectations increased. So recovery takes longer at 2% inflation than it would at 3-4% inflation. The result of "The Conquest of American Inflation" has been now three jobless recoveries.

Larry Hamelin said...

Ken, You're mostly correct, but I think you have the terminology backwards: the real rate (r) is the nominal rate (i) - expected inflation (pi^e). But your formulation is still essentially correct, just backwards from how it's usually presented in undergraduate economics classes.