There’s a reason it’s called “monetary” policy
Although I am not really a monetarist, I do like their basic approach, especially the way they distinguish between money and credit. Changes in the price level are determined by changes in the supply and demand for the medium of account, i.e. money. Unfortunately, the Fed seems to have forgotten that, and increasingly seems to focus on credit. Here is Dave Altig of the Atlanta Fed:
Is the 25 basis point return paid by the central bank creating a significant incentive for banks to sit on reserves rather than lend them out to consumers or businesses? At least some observers are skeptical:
“Barclays Capital’s Joseph Abate…noted much of the money that constitutes this giant pile of reserves is ‘precautionary liquidity.’ If banks didn’t get interest from the Fed they would shift those funds into short-term, low-risk markets such as the repo, Treasury bill and agency discount note markets, where the funds are readily accessible in case of need. Put another way, Abate doesn’t see this money getting tied up in bank loans or the other activities that would help increase credit, in turn boosting overall economic momentum.”
This is one of those situations where I don’t know whether I should be criticizing Abate, for what he said, or Altig, for seeming to approve of it. But whoever is to blame, this quotation seems to get things exactly backward, and does so by focusing on credit, rather than money.
Let’s suppose banks reacted to a cut in the interest rate on reserves (IOR), by switching over to some sort of short term debt. Where would the reserves go? There is only one place unwanted bank reserves can go, out into circulation. That means a nearly instantaneous $1 trillion increase in currency in circulation, with banks instead holding more short term assets like T-bills.
[As an aside, I do understand that this probably wouldn’t happen, I’m just trying to work through Abate’s worst case. In practice, yields on T-bills might fall to zero, or even a few basis points below zero. Of course if there was a negative interest rate on ERs (including vault cash), then it really would happen.]
So what would this mean for the economy? First, it would be equivalent to a trillion dollars in quantitative easing (QE), which is quite a bit. But you might ask; “Why would an extra trillion in QE help, didn’t the last round of QE have relatively little effect?” One answer is that at least it had some effect, it was better than nothing. But more importantly, the previous QE had little effect because most of the new base money went into ERs. If the banks got rid of reserves and replaced them with short term debt, then all the extra base money would go into circulation. It would be like a trillion in QE, with 100% of the new base money becoming currency in circulation. Not just QE, but super-charged QE. Indeed this is the nightmare scenario that led to the IOR program in the first place. The Fed was doubling the monetary base in late 2008, and feared what would happen if all that extra cash went out into circulation. What Abate defines as failure, to me seems like a dream come true.
Of course a super-pessimist like Paul Krugman might talk about how sales of safes boomed in Japan once rate hits zero. And technically it is possible that even this super QE would have no effect—it might all be hoarded. But a trillion dollars is a lot of cash, and I think it more likely that the “hot potato” process would take over—people would try to spend some of the extra money, driving up AD.
The “hot potato” process is the central concept of monetary economics. If you put more cash into circulation than people want to hold, they’ll try to get rid of it. Individually they can, but collectively they cannot. The attempt to get rid of the cash will drive up AD. I think everyone understands this (excluding post-Keynesians of course) but many economists forget what monetary policy is all about; the supply and demand for money. Lowering the IOR will lower the demand for money, and tend to raise prices and NGDP. The effect may be small or large, it mostly depends what else the Fed does. But with the economy this weak it is certainly worth trying.
HT: Mark Thoma, JKH, JimP
Tags:
1. August 2010 at 07:52
Safes and Gold.
Instead we could raise rates and sell off foreclosed assets…. which would generate HUGE activity far past $1T in hoped for Treasury QE.
Selling off the underlying assets at a loss is a GIANT POSITIVE, that is an admittance that GSE+ MBS = BS – which is morally healthy, and again:
1. Drastically improves balance sheets of the guys we want to INVEST. The $2T in dry powder guys sitting on sidelines.
2. Reduces rents for lower and middle class home buyers, so they can buy elsewhere.
3. Drives the zombie insolvent banks out of business.
4. Lets every underwater home buyer make the final decision once and for all. Add in mobilizes the workforce.
——
Just sat and listened to Alan “banks first” Greenspan say we have to worry about Home Prices.
Screw the banks. When dry capital is jumping for joy, everything else is going to be fine.
1. August 2010 at 09:07
Jim Hamilton agrees – or at least Andy Harless does in the comments.
http://www.google.com/reader/view/#stream/user%2F15470056746793432676%2Flabel%2FFinance
Fisher wishes to crush us – for the pure fun of it.
http://yglesias.thinkprogress.org/2010/08/dallas-fed-pushing-zero-inflation-agenda/?utm_source=feedburner&utm_medium=feed&utm_campaign=Feed:+matthewyglesias+(Matthew+Yglesias)
I find it amazing that Obama, or at least Summers, seems to be entirely unaware of this debate.
Maybe it really is true that Summers and Bernanke will not change policy now – out of the fear that it might work and then they will be asked why they did not do so a year ago.
But surely Obama does not actually want to lose the election. That could not possibly be true.
1. August 2010 at 09:48
If the Fed is worried about the effect on unleashing the whole Trillion on the market, maybe they could try a more sequential approach. They could put a large cap on the amount of excess reserves eligible for the interest – say, $100 Billion per institution (or, maybe, to start, just $10 Billion below the amount held by the largest ER-holding bank). That $10 Billion gets put into the system, and they have a little time to sit back and watch, cautiously and prudentially, what happens.
Then the next week, they lower the cap another $10B (or by whatever amount that puts an aggregate of another $10B into circulation.) If things are going well, they can accelerate a little, if inflation gets “out of control”, they can pull back.
It would be like having another little precision lever to push when you’re “stuck” at the zero bound. A way that allows you to still press harder or softer on the throttle when you’re already stuck in fifth gear.
You could it the “Reserves Precision Tuning Channel” or something jargon-esque like that.
It would help save face too. No one has to admit a change in policy or an error in intellectual understanding, and it’s essentially conservative in that “try just a little first, wait, assess and evaluate, and then carry on if results are acceptable” kind of way.
Any thoughts?
1. August 2010 at 10:13
The hot potato process only works when the potato is actually hot, which is to say, when safe, short-term investment vehicles pay interest, so that there is a disadvantage to holding more money than you need. If T-bills pay zero (or negative) interest, there is little reason to cut down ones holdings of currency: you may have no anticipated need for it as a medium of exchange, but it’s a perfectly good store of value. Issuing more currency is just like issuing more T-bills. (Granted, there might be costs and risks associated with storage of cash, but these are small and would be reflected in a slightly negative T-bill yield.)
If you want to make a hot potato argument with zero- or slightly-negative-yield T-bills, you have to include those T-bills among the hot potatoes. If the banks get rid of their reserves by buying T-bills from the public, it doesn’t change the number of hot potatoes. To successfully increase the number of hot potatoes, the Fed (or someone) needs to buy longer-term (or riskier) securities, thereby making them “hot” (i.e. brining their yield down to near zero).
1. August 2010 at 10:30
Perhaps like some other readers, I find myself struggling to fully grasp monetary policy and Sumner’s takes on event-policies, although I really enjoy what I do understand.
Scott–is there a basic primer I could read, something (I can hope) brief, in the 50-page range or less?
I have a good understanding of economics, took plenty in college, was a financial journalists, but when you get professional economists on different sides of a monetary policy question, I tend to get lost.
1. August 2010 at 11:48
If the interest rate on some asset is close to zero, then it and money are near perfect substitutes. In this case, expansionary monetary policy can only be pursued by purchasing assets with higher interest rates. That is what the Fed must do, so what is the problem? I don’t get it.
At some point, the demand for money will be satiated and people will begin spending. However scary it might be to see the money supply increase by such a large quantity, it is merely in the pursuit of monetary equilibrium. Surely economists like equilibrium, right?
1. August 2010 at 14:56
Morgan, Duly noted.
JimP, I couldn’t open the first link. Thanks for the second.
Indy, Last year I discussed a similar idea. Three layers of reserves. RRs, interest bearing non-required, and beyond some threshold, non-interest bearing ER.
Andy. Then why is current holdings of currency only about $850 billion, which is not much more than before the recession. If near zero rates made people want to hoard a lot of cash, then why aren’t they doing it already? People are free to adjust their currency holdings, and don’t seem to want to hold all that much more currency than when rates were far higher. So I’m confident there would be at least some of the hot potato process. Indeed so much that we’d overshoot toward excessively high inflation. Hence I think the Fed would actually pull some of the currency out of circulation, rather than allow currency in circulation to jump to 1.85 trillion. I think fear of theft is a non-trivial concern.
BTW, I do not believe pouring a trillion in currency is a sensible policy, just saying it would probably “work.”
Benjamin, I often get that question, and don’t have a single good answer. But I did list some readings at the end of the FAQs on the right side of my blog. It’s around question 20, I believe. You can check the list out.
Lee, See my answer to Andy. Cash and T-bills aren’t even close to being perfect substitutes at zero rates. Most cash is held in the underground economy. They can’t substitute T-bills for cash at low rates, because they aren’t holding any T-bills right now. You can’t use T-bills to go shopping. There’s no risk of theft for T-bills. There is a huge difference between the sort of institutions that hold T-bills and the individuals that hold cash. Sure, if deflation got bad enough some conventional investors might stuff cash in safety deposit boxes, but we aren’t even close to that situation, indeed inflation expectations would surely rise if the Fed did something aggressive, so it would not be a problem in the real world, no matter how worrisome it seems in textbook models.
1. August 2010 at 16:41
Benjamin,
Svensson’s Monetary Policy with a Zero Interest Rate might be a good start (only 14 pages): http://people.su.se/~leosven/papers/MonPolZIR090217e.pdf
1. August 2010 at 18:23
The reason people haven’t increased their holdings of currency is that they can keep the money in a free — or nearly free — federally insured bank account. If the banks refused to accept that cash as deposits, or started charging substantial proportional fees on bank accounts, people would start holding more of their money as cash instead.
It doesn’t make any sense to me that people would “try to spend some more of the extra money.” Cash is an asset, just as a bank account is an asset. People don’t change their spending habits just because of a shift in their portfolio. There might be a very tiny increase in spending, due to the reduction in the return on their portfolio (i.e. the cost of insuring, or shadow cost of not insuring, their cash), which might (assuming that the substitution effect predominates over the income effect) make current consumption slightly more attractive relative to future consumption. But the effect would be very tiny, just as if the interest rate on your savings account were reduced by 25 basis points.
Instead, what would happen to the “hot potato” is that people would readjust their portfolios by attempting to buy other assets, until the yields on other assets had come down enough to make people indifferent, on the margin, between holding those assets and holding cash. The yields wouldn’t have to come down very far. It’s difficult to estimate how large the effect would be, but on any given asset, it would be somewhere between zero and 25 basis points. There’s no way that a 25 basis point drop in IOR could result in a larger than 25 basis points drop in the required yield on some other asset. I’m guessing the effect would be closer to zero, but that’s just a guess.
A 25 basis point drop in yields might be enough some effect on the economy, but it would not be a large effect, even in the most optimistic case. If my guess is right and the drop is closer to zero, then the effect would be even smaller.
2. August 2010 at 02:55
Scott,
In anticipation of my own ignorance regarding T-bills, I didn’t actually mention them in my previous post. In any case, by “money” I had in mind both currency and bank accounts. For my response, see the first paragraph of Andy Harlass’s last post.
This brings me to another matter. Does the FDIC make money demand more volatile by socialising the risks of holding money?
2. August 2010 at 03:08
not relevant to this topic, but here’s an interesting post:
http://economicsofcontempt.blogspot.com/2010/08/why-fed-isnt-doing-more.html
2. August 2010 at 04:26
Thanks Benjamin,
Andy, I have discussed eliminating interest on ERs only. Often I forget to mention that. Since the Fed could continue to pay interest on required reserves, there would be less of a need for negative interest rates on bank deposits–although it still might happen.
But I do understand your basic argument, it is possible that interest rates would fall so far that people stuffed cash into safe deposit boxes. But interest rates falling “so far” is supposed to be a good thing, when trying to boost AD, isn’t it? My point is that it would take interest rates well below a few basis point below zero, to get the public to hoard an extra trillion in cash. BTW, I think that nominal interest rates would not fall, instead inflation expectations would rise–reducing real rates.
You said;
“It doesn’t make any sense to me that people would “try to spend some more of the extra money.” Cash is an asset, just as a bank account is an asset. People don’t change their spending habits just because of a shift in their portfolio. There might be a very tiny increase in spending, due to the reduction in the return on their portfolio (i.e. the cost of insuring, or shadow cost of not insuring, their cash), which might (assuming that the substitution effect predominates over the income effect) make current consumption slightly more attractive relative to future consumption. But the effect would be very tiny, just as if the interest rate on your savings account were reduced by 25 basis points.”
I’m afraid I have to disagree here. You are criticizing the “hot potato” transmission mechanism, you are doing so by assuming a Keynesian transmission mechanism. But by this logic, a cash infusion would have little or no effect on AD even if wages and prices were perfectly flexible. Consider a doubling of the supply of currency in a world where wages and prices are completely flexible. The price level immediate doubles, and interest rates are unchanged. But if we assume (as Keynesians do) that the extra AD came from the people spending more because they feel richer, then AD should not rise at all. After all, if you swap T-bills for cash, people don’t feel any richer due to that effect. And by assumption interest rates don’t fall (wage and price flexibility.) So how does NGDP double during a period when interest rates are well above zero and wages and prices are completely flexible? The answer is the hot potato effect–people try to get rid of a specific asset. By analogy, if their is a huge apple crop, and the relative price of apples falls in half, then nominal GDP in apple terms doubles. To explain that effect you do not need to resort to some effect based consumers feeling richer and buying more goods and services.
Sorry for being so long-winded, but it seems to seems to me that you have criticized the hot potato effect at the zero bound, by simply assuming an argument that implies it doesn’t exist even in normal times.
I do agree with you on one point, if we are going to assume that monetary policy works through lower nominal interest rates, then I agree that monetary policy is an exceedingly weak tool, indeed it would be exceedingly weak even during normal times, as easy money usually doesn’t lower long term rates. But the “hot potato” transmission mechanism (which I think all mainstream economists accept as a long run proposition), doesn’t work through interest rates, as interest rates are unaffected in the long run by one time changes in the money supply.
Lee, Because of the very large money multiplier in modern banking, if the money went into banks (as required reserves) then the broader monetary aggregates would rise sharply–which is supposedly what we want.
Thanks q, I hope politics isn’t behind the Fed’s stubbornness
2. August 2010 at 05:23
Prof. Sumner, could you comment briefly on what the specific factors were that led to the increase in demand for money back in 2008? I’m still trying to get it straight in my mind what the causes and effects were for tight money. Thank you.
2. August 2010 at 09:07
Malavel-
thanks for the tip. I will read (and maybe even understand!).
2. August 2010 at 10:33
I’m not assuming a Keynesian transmission mechanism; I’m acknowledging it as a possibility, which I reject as being inconsequential. My argument becomes stronger if you assume there is no wealth effect.
My point is that money is not a unique asset. An increase in the price of goods in terms of money is also an increase in the price of goods in terms of bonds, since bonds are denominated in money. If banks get rid of their reserves by buying T-bills from the public, they have decreased the supply of T-bills (“bonds”) to the public. Why don’t people then try to sell all their goods (and drive down the price of goods) so that they can own more of these now precious T-bills?
The answer is that the resulting equilibrium price of goods depends on the relative substitutability and complementarity among money, bonds, and goods. Money and goods are more complementary than are bonds and goods, and that’s why the price of goods goes up instead of down. But if money and bonds (in this case, T-bills) are very close substitutes, then goods prices in the new equilibrium will be almost the same as goods prices in the old equilibrium.
In the monetarist model with exogenous velocity, there is zero substitutability between bonds and money, so the effect of the increase in money is fully reflected in the price of goods. At the other extreme, one could assume that money and bonds are perfect substitutes, in which case money (when it displaces bonds) has no effect on the price of goods. The reality today is much closer to the latter.
2. August 2010 at 10:38
Scott re: Andy and hot potatoes,
I think your apple analogy is incomplete. What if the crop of delicious apples doubled at the expense of half the crop of Granny Smiths? If they are close enough substitutes, Delicious prices don’t drop by 50%, and may not even move much. I don’t think it’s about whether prices are flexible, but whether the zero bound prevents T-bills from dominating money as a store of value. Since the Fed won’t try sub 0% rates to find out the cost of holding paper dollars, this zero bound applies to deposits as well in practice.
You usual argument is that if the new money is thought to be permanent, then people will expect money to eventually lose its portfolio value because eventually rates will rise and T-bills will dominate. That itself would raise inflation expectations, but wouldn’t be well described as a hot potato effect in the first order. First, rates would rise on expectations of future inflation, and then the hot potato effect would kick in as the demand for money as a store of value would decline as a spread opened between money and T-bills.
2. August 2010 at 10:53
@Andy,
If excess reserves paying no interest are not hot potatos, then the Fed can just buy up all existing government debt, and no one will mind. Not only that, Obama can cut taxes to zero and just let the Fed buy all the resulting T-bills. If you think such a policy would be insane and result in massive inflation, then you don’t actually believe that excess reserves and Treasury debt are very close substitutes.
2. August 2010 at 11:50
@Jeff
“If excess reserves paying no interest are not hot potatos, then the Fed can just buy up all existing government debt, and no one will mind.”
Fed purchases of short-term government debt (which is what we are contemplating when we talk about banks) will make negligible difference, in my opinion, even if the Fed buys up nearly the entire outstanding stock of such debt, which would be much more than the reserve liquidations this post contemplates. (Purchases of long-term debt are another story. They will make a difference by lowering long-term interest rates, because long-term bonds are not such a close substitute for money. I’m’ not sure how big the effect would be. Depending on how they are perceived by the market, Fed Treasury purchases could make a difference by affecting expectations (but that’s actually an irrational response by the market, unless the Fed is intentionally signalling its willingness to allow higher inflation). The expectation effect has nothing to do with asset substitutability; it’s all about signalling.)
“Not only that, Obama can cut taxes to zero and just let the Fed buy all the resulting T-bills.”
No, that is a completely different policy, because in that case, the money is not displacing bills that are already in the market. (The Treasury would have to issue new T-bills for the Fed to buy, so the supply of “money plus T-bills” goes up.) By contrast, if banks replaced their reserves with T-bills, they would have to buy the T-bills from the public, which would reduce the supply of T-bills exactly as much as it increases the supply of cash. That’s my whole point.
3. August 2010 at 05:10
Tim, The main factor was probably the banking crisis. Money is the most liquid asset in a crisis. But once inflation expectations began to fall, that also increased money demand.
Andy, You said;
“My point is that money is not a unique asset. An increase in the price of goods in terms of money is also an increase in the price of goods in terms of bonds, since bonds are denominated in money.”
The term “denominated” is a bit vague here, unless you are assuming the price of T-bills is fixed because we are in a liquidity trap. If bond prices can change, then they are no more a medium of account that copper or zinc. So I will proceed under the assumption you are restricting your argument to zero rate T-bills.
You said;
“In the monetarist model with exogenous velocity, there is zero substitutability between bonds and money, so the effect of the increase in money is fully reflected in the price of goods. At the other extreme, one could assume that money and bonds are perfect substitutes, in which case money (when it displaces bonds) has no effect on the price of goods. The reality today is much closer to the latter.”
I’d like to see some empirical evidence for this. It seems that many economists simple look at the rate of return on cash and T-bills, see they are roughly equal, and assume they are close substitutes. But I just don’t see that at all. Rate of return is not the only relevant comparison. You can’t assume the answer, you must show it. And I think most empirical evidence suggests cash and T-bills aren’t close substitutes in the US:
1. The demand for cash has not risen sharply as nominal rates have fallen close to zero.
2. Cash and T-bills are not at all close substitutes for cash hoarders, because cash hoarders are motivated by secrecy. Even when we are not in a liquidity trap cash hoarders comprises 75% to 80% of cash demand.
3. Cash and T-bills are also not close substitutes for the 25% of cash in transactions balances. T-bills are not a convenient way to go shopping.
4. You ignore expectations effects. The monetarist argument does not rely on the two assets not being close substitutes today, rather it relies on money and bonds not being close substitutes in the long run. And that is a very reasonable assumption, as liquidity traps don’t last forever.
dlr, I understand the close substitute argument, but Andy was criticizing the hot potato argument with an example that didn’t rely on T-bill yields being near zero. He was talking about the lack of a first order wealth effect when you inject cash into the economy by swapping bonds for cash.
If we are going to discuss close substitutes, that’s fine, but that is a very different argument.
Regarding your second paragraph, my long run argument is a long run hot potato argument. I’m saying the hot potato effect is expected to work in the long run, that raises inflation expectations, which makes them a hot potato right now.
But I am also saying that they aren’t perfect substitutes even today, so you get somewhat of a hot potato effect immediately, even w/o expectations effects.
Andy, I mostly agree with your reply to Jeff, and would go a bit further. The expectations theory of the term structure of interest rates suggests that even the expected 3 month rate on long term bonds is near zero. So in that case even buying long term bonds wouldn’t have much expansionary effect via Keynesian channels. Of course I think it would have effects through other channels (if not sterilized with IOR) but that is another argument.
Krugman makes this argument about the reductio ad absurdum case of the Fed buying up massive amounts of all sorts of assets all over the world. He basically says “Yes, it would work, but it would expose the Fed to too much investment risk when it later sold off part of that portfolio to prevent hyperinflation.”
3. August 2010 at 08:50
It’s my hypothesis that the “factions” at the Fed theoretically hindering Bernanke from pursuing an explicit inflation target, doing more QE, etc. are a result of the “supply-side” ideology. Here are the pillars of supply-side economists:
1. Low marginal tax rates (and gov’t spending)– this is the more famous pillar.
2. A central bank that pursues 0% inflation, Volcker being the staunchest advocate of this. Remember last year at a forum he grilled a Fed governor about “why not just pursue 0%?” Every supply-side blogger and economist I know of abhors Dr. Sumner’s idea of NGDP targeting, and abhor the idea of an explicit inflation target other than 0%. Even a consistent level of 2% is anathema.
Look at Fisher in Dallas, Hoenig in KC, etc. All fall under this idea. They somehow reject the idea that a 0% target creates a greater danger of a liquidity trap.
3. A strong-dollar policy. Every supply-sider I know is decrying a depreciating U.S. dollar, and they blame Bernanke and Obama’s policies for it.
In my view, Fisher, Koenig, etc. and a wide range of bloggers (Rajan included) fall into this camp. Hence, they want the Fed to pursue a tight money policy to keep appreciate the dollar and LOWER inflation to 0%. Their justification is that “inflation is the ultimate evil for an economy.” Krugman is right, they don’t see deflation as a bad thing either.
Is my hypothesis wrong? If so, please enlighten me.
3. August 2010 at 13:05
Scott,
What matters is substitutability at the margin. Your examples are inframarginal. Presumably the demand for cash hasn’t risen because cash and bank deposits are, at the margin, nearly perfect substitutes. The smugglers and vending machine users have all the cash they need, and they already had all the cash they needed, or they would have been holding more cash, and fewer bank deposits, in the first place.
In order for the original example to make sense, it must assume that cash and bank deposits are close substitutes; otherwise those erstwhile excess reserves couldn’t get into circulation in the first place. Given that banks don’t currently charge proportional fees on deposits, a 25 bp cut in IOR wouldn’t lead them to charge any more than 25 bps, so the assumption must be that 25 bps would be enough to get people to hold all that new cash as cash rather than depositing it.
So we haven’t really observed the degree of substitutability between cash and bonds, because bank deposits are closer substitutes for both than either are for each other. If we’re trying to imagine a world where a lot of cash suddenly appears and doesn’t get immediately converted into bank deposits, we have no data on which to base our estimates. All we can do is make analytical and intuitive arguments.
As for the long run,
1. What does the current IOR rate have to do with the long run? Surely everyone would expect any suspension of IOR to be temporary. Fed officials have explicitly mentioned IOR as part of their exit strategy.
2. I’m not sure cash and bonds are better substitutes in the short run than in the long run. Based on observing my own behavior, I would say that household asset allocation choices respond to changes in interest rates but do so slowly. Most demand curves are more elastic in the long run than in the short run. Should the opposite be true of demand for cash?
3. In the long run (i.e. at maturity), bonds automatically convert themselves into cash, so I would expect the two to be fairly close substitutes. The main advantages (liquidity, and inflation protection — since cash can be invested later at a potentially higher nominal interest rate) are on the side of cash. At the margin where cash is no longer scarce, I would expect the two to be nearly perfect substitutes. (I’m thinking that bonds and cash are like fresh cider and hard cider, except that hard cider hyperinflates into vinegar, and many people prefer drink fresh cider before it goes hard, so the analogy doesn’t work; but still, the two are substitutes.)
3. August 2010 at 19:10
JTapp, I don’t know. I’ve had commenters come on here and insist that supply-siders liked easy money. They cited George Gilder. If Volcker says he favors zero inflation then he did a lousy job as Fed chairman. Inflation was a pretty stable 4% per year from 1982 through 1987. (excluding an oil price dip in 1986). So he acted like 4% was his target. But my hunch is that you are mostly right.
Andy, Interestingly, I had a long discussion last year on this exact topic with people like Bill Woolsey. I’m told teh M2 multiplier is huge, something like 80. So there is no real likelhood of the trillion in base money going into the banking system as required reserves. That would produce $80 billion in M2. Thus if you assume that banks don’t want to hold lots of ERs (which was the Abate assumption I worked with), then the cash goes out into circulation. At this point the debate becomes very hard to define, because almost anything is possible depending on expectations. You can definitely envision a scenario where all that cash is stuffed in safes, I just don’t find it very plausible. I think a more likely outcome is that AD would start rising dangerously fast, and the Fed would be forced to pull lots of money out of circulation, to prevent overheating. I think that in order to get people to hold that much cash, you have to have interest rates go significantly negative, not just on bank deposits, but also T-bills. And I just don’t see that as likely, especially if this all occurred very rapidly and as part of an explicit Fed policy of easing. Is it theoretically possible? Sure. But not likely.
You said;
“If we’re trying to imagine a world where a lot of cash suddenly appears and doesn’t get immediately converted into bank deposits, we have no data on which to base our estimates. All we can do is make analytical and intuitive arguments.”
My intuition is that this sudden flood of currency would produce absolute chaos, and inflation expectations would soar.
You said;
“1. What does the current IOR rate have to do with the long run? Surely everyone would expect any suspension of IOR to be temporary. Fed officials have explicitly mentioned IOR as part of their exit strategy.”
It drove me crazy when we had all these Fed people saying “don’t worry, the currency injections are just temporary, we will sterilize them or perhaps pull the money out of circulation in the future.” That is exactly the sort of temporary currency injection that Krugman (correctly) pointed out was ineffective. But the Fed obviously cannot commit to a future money supply or specific IOR, so what they need to do is have some sort of explicit nominal target. If they don’t, then monetary policy becomes much more difficult. I have often argued that even massive QE is ineffective if expected to be temporary. And that’s even true at zero rates. So the entire IOR/QE approach is extremely clumsy, it is essentially trying to indirectly create inflation expectations, instead of just coming right out and doing it explicitly.
The reason I thought the Abate example was likely to “work” is that it probably would raise inflation expectations, given that it would be seen as a fairly reckless and extreme policy. A trillion in ERs in no big deal–they are like little bonds. A trillion in actual currency would certainly get the public’s attention.
You said;
“2. I’m not sure cash and bonds are better substitutes in the short run than in the long run. Based on observing my own behavior, I would say that household asset allocation choices respond to changes in interest rates but do so slowly. Most demand curves are more elastic in the long run than in the short run. Should the opposite be true of demand for cash?”
Good point. I just meant that rates on T-bills were likely to be well above zero in the long run. But you are right, it is easier to adjust currency holdings in the long run, for any given interest rate. But if T-bill yields were 3% 10 years from now, then then would not then be close substitutes for cash as a store of value.
On your third point, I think we are back into the zero rate situation, where cash and T-bills are close substitutes as a store of value. But again, I don’t expect that situation to persist for long (unless the Fed continues with its foolish policy of being passive.)
I will be traveling, but will try to return to this in a few days if you have further comments.
BTW, I didn’t comment on your structual unemployment post, but thought it was quite good.