What’s this “Tinkerbell” stuff all about?
It occurred to me that perhaps much of what I have been discussing recently is a bit too esoteric for normal people who don’t live and breathe Woodfordian monetary theory. So today I’m going to try to explain the basic ideas in a very simple way. Then in part 2. I’ll try to explain how I can use the same Woodfordian model that people like Thoma and Krugman use, and reach different conclusions.
I’ll start where Nick Rowe left off yesterday. Nick spent a lot of time discussing all the perplexities of trying to control the economy by controlling real interest rates. Unfortunately my brain is not wired properly to understand monetary policy based on manipulating real interest rates. I see the new Keynesians as taking a peripheral stylized fact (prices are sticky), exaggerating to the point of inaccuracy (prices don’t change at all in the short run), and then making it centerpiece of their model. But Nick ends on a more hopeful note, which is where I’ll pick up:
Tinkerbell and framing aside, this reveals another critique of our thinking about monetary policy as setting interest rates. Why did we ever think that cutting real interest rates would increase demand permanently, as Old Keynesian models suggest? Cutting real interest rates merely shifts demand towards the present, and away from the future. That won’t work if both present and future demand are too low. Maybe monetary policy is about the supply and demand for money?
I’ll start with the last line of Nick’s post. The only knowledge I am going to assume that you have is an understanding of the Quantity Theory of Money—the idea that if you double the money supply, the price level will also double over time, leaving every real variable in the economy unchanged.
I want you to imagine that everyone understands and believes in the QTM. Imagine you live in a country where a typical 3 bedroom ranch house sells for $200,000. Also assume the money supply has been stable for years. Now the Fed suddenly doubles the money supply. What will happen to the price of that house? Keynesians will say “nothing”; prices are sticky. If they are right, I plan to buy up as many houses as I can, right after the money supply doubles. And then sell them again when the house prices double later on. But I actually think it more likely that the sellers will also understand this implication of the doubled money supply, and won’t hand me a $200,000 profit on a silver platter. They’ll immediately demand higher prices. The Keynesians are right that in the real world many prices rise more slowly, but in any case they do eventually rise.
So far there is nothing controversial in my application of the QTM. But now let’s assume that as the Fed doubles the money supply, they announce that 4 months later they plan to withdraw the extra money from circulation. Does the price level still double in that case? Or, if you are a sticky-price Keynesian, does the expected future price level still double? If we think about that $200,000 house, I think the answer is clear. Who in their right mind would pay $400,000 for a house expected to be worth only $200,000 very soon, after the monetary injection is withdrawn in 4 months? Indeed we would expect almost no increase in the price of the house, despite the doubling of the money supply. And the reason is simple, the efficient markets hypothesis is far more fundamental than the QTM. It’s simply not plausible that house prices would rise from $200,000 to $400,000, if people expected them to return back to $200,000 in the near future.
So let’s review. If you double the money supply, and the increase is expected to be permanent, speculators will rapidly bid up the prices of houses. And even the prices of sticky goods will be expected to rise as their prices are readjusted over time. But if you double the money supply, and the monetary injections are expected to be withdrawn in the near future, then house prices will barely budge, and sticky prices won’t be expected to eventually rise upward. This is an incredibly powerful insight.
What does this all mean? It means that the effect of changes in current monetary policy on current aggregate demand and prices is utterly trivial compared to the effect of changes in the future path of monetary policy on AD and prices. Change the current money supply and leave all future money supplies unchanged, and almost nothing happens. Leave the current money supply unchanged and change all future money supplies for year 2, 3, 4, etc, and you have a powerful and immediate effect on AD and prices. The current price level basically depends on the future expected path of monetary policy. Woodford didn’t discover this idea, but it forms the centerpiece of his model.
OK, but how does all this relate to the “Tinkerbell principle?” And why do we only hear about this principle during “liquidity traps?” The reasons are complicated. One reason is that Keynesians like Woodford and Krugman think of monetary policy in terms of the path of interest rates. This leads to very different assumptions from using the money supply as your policy instrument.
Let’s redo the previous example with interest rates. The Fed cuts the interest rate this year, but leaves all future expected interest rates unchanged. Unlike with the monetary base, the expansionary effect of today’s action is not completely negated by the fact that future monetary policy is unchanged. Those who trust me can skip the next paragraph.
[You can think of this in two ways. Assume the current expansionary cut in rates boosts the economy this year. Because AD is higher at the end of the year, even an unchanged level of future interest rates means an effectively lower policy rate in the Wicksellian sense. The rate has fallen relative to the natural right (which is higher in a stronger economy.) If this makes no sense here is another explanation. To cut rates this year the Fed must increase the monetary base. This boosts NGDP. If they plan to keep all future rates unchanged, then they must keep a higher monetary base at the end of the year when interest rates return to their normal level. This is because with more spending in the economy, you have more demand for base money, and hence must supply more base money to keep rates at the predetermined level. So a one-time cut in short term rates leads to a permanent rise in the monetary base. That’s why even temporary monetary policy changes can have an effect if you use interest rates as your policy tool. but even there, future expected changes are far more powerful.]
What about the “liquidity trap?” Recall that Keynesians think that monetary policy becomes ineffective once rates hit zero. (Modern Keynesians use short term rates, but Michael Belongia reminded me that Keynes actually was thinking in terms of long rates. So we aren’t actually in a Keynesian liquidity trap, as the Fed’s QE in March 2009 did substantially affect long rates.) You might wonder; “Why don’t the Keynesians think a permanent doubling of the monetary base would raise prices, even if interest rates were zero today?” After all, my initial example with the $200,000 house and doubled money supply seems very straightforward. The answer is that smart Keynesians like Woodford and Krugman and Thoma do understand that a permanent doubling of the money supply would raise prices. Their argument is different from Keynes’s original liquidity trap argument. They fear that any monetary injection would not be expected to be permanent. More specifically, they fear that the Fed would not be able to convince the public that the increase would be permanent. And if they can’t do that, then they can’t convince them that the price rise would be permanent. And if that $200,000 house is only expected to temporarily rise to $400,000, then it will never rise in price in the first place. So to create current inflation you have to believe the inflation will be permanent. “If we believe we can inflate, then we can inflate.” That’s Tinkerbell.
Part 2.
So far I agree, but now let’s look where I disagree with the standard Woodfordian approach to monetary policy traps. To do that we need to think more about what the Fed is really doing. Recall that the most important thing the Fed does is not to set the current value of the money supply (or interest rates), but rather to signal intentions about future policy. But how do they do this? There are many ways. They could change the monetary base, and let the public guess what that meant. They could announce permanent changes in the money supply. They could adjust the exchange rate. They could announce that they are targeting the expected inflation rate in the TIPS markets. And so on. In practice, most central banks send signals by changing short term nominal interest rates.
Unlike all the other options that I mentioned, nominal interest rates have an Achilles heel. They might need to go significantly below zero, but cannot. This means that if rates fall to zero, and the Fed wants them to be lower, and the Fed is incapable of communicating with the public in any way other than interest rate changes, then the Fed becomes literally dumb (in the sense of speechless, although I’d argue that slang for ‘stupid’ also applies here.)
[Note, here and in a few other places I am shamelessly stealing ideas from Nick Rowe’s brilliant “social construction of monetary policy” post.]
So the markets look to the Fed for direction, and they have nothing to say. During the first 10 days of October, 2008, the markets saw that the short term rate needed to go well below zero to prevent a severe recession. They looked to the Fed for some signal that it was switching out of interest rate “talk” and into some other language like price level targeting. But like most big bureaucracies, the Fed is not as nimble and quick as Tinkerbell. They remained mute—and the asset markets understood that this meant future monetary policy would be constrained by the zero rate bound on short rates, and would be far too contractionary. Asset prices crashed.
Modern central banks don’t just have one language, they have two. In addition to signaling short term intentions with interest rates, they signal long term policy goals with inflation targets. So even when rates hit zero, the Fed could have signaled a higher inflation target. Indeed Krugman and Woodford both recommended this. This would be a signal that once we exited the liquidity trap, the Fed would keep monetary policy more expansionary than usual, so that prices could rise by more than the normal 2%. Here’s where the expectations trap comes in. Once we have exited the liquidity trap, the Fed might not want the high inflation to actually occur. It makes sense to promise inflation, as that will lower real interest rates today and help us recover. But once the deflation and recession have ended the Fed might renege on this promise, because they are conservative central bankers who don’t like high inflation.
Here’s an analogy. A mom promises a child that she’ll get a lollipop if she finishes her homework. After homework is finished, the mom reneges on her promise, because lollipops are bad for the child’s teeth. And after all, the homework is done so the inducement has achieved its purpose–even if it was all a lie. I hope you can see the problem here. This sort of thing almost never happens. Moms do give the lollipop, and for two very good reasons:
1. Moms are not evil witches.
2. Moms may have to promise lollipops in the future.
So although the “expectations trap” is a nice clever theory, it almost certainly has no implications for the real world. If the central bank publicly promised a certain inflation or price level path in an emergency, they would almost certainly carry through with the promise. Why then was this silly theory developed? Because I think people gave far too much respect to the Bank of Japan protestations of innocence in the 1990s, and focused far too little on the fact that the BOJ was unwilling to actually promise inflation. So it looked like the BOJ was stuck, unable to move AD and prices, when in fact they weren’t really trying. Even worse, the BOJ actually did eventually do a temporary currency injection somewhere around 2001-02, and then withdrew the money in 2006. And remember, the whole expectations trap idea is based on the insight that temporary currency injections have almost no effect. So when the temporary currency injection had almost no effect, it seemed to support the model. It does support the model that temporary currency injections don’t have much effect, but doesn’t support the assumption that the central bank can’t signal inflation. After all, they never tried to signal inflation.
So why does the Woodford model seem to suggest that fiscal policy can work when monetary policy is stuck in an expectations trap? This is very complicated, as (I am pretty sure) it rests on two dubious assumptions:
1. The Fed can only signal short term policy by targeting short term rates.
2. The Fed targets inflation at 2%, come hell or high water.
If you buy both assumptions, then it is indeed true that when short term rates are zero, the Fed can do nothing. They can’t cut short rates and they won’t change their long run 2% inflation target. And it’s also assumed that if they did promise higher inflation, no one would believe them.
Here’s my problem with that view. The term ‘trap’ suggests there is nothing they can do. But in fact this “trap” is completely self-inflicted. They can target some other variable, such as the money supply, or long term interest rates, or exchange rates, or TIPs spreads, and signal a more expansionary policy in that way. They aren’t dumb (well, at least in the “mute” sense.) And even Woodford himself has argued that if they are at the zero rate bound, their long run target should be the price level, not inflation. Either changing the short run policy indicator or changing the long run policy goal would allow them to boost AD.
A skeptic might say: “But you haven’t really addressed the expectations trap. Suppose they set a higher price level target, and no one believes them? Then we are still stuck.” I have three problems with this view:
1. They would be believed.
2. Even if they weren’t they can use other short term tools like currency depreciation or TIPS spread targeting.
3. If there is still a problem it is even more applicable to fiscal policy.
And now we have come full circle to my post that raised such a fuss. I argued that if you take the Woodfordian view that future expected monetary policy has a far greater impact than changes in current monetary policy (which I accept) then it is equally true that changes in future expected monetary policy can have far more effect than changes in the current stance of fiscal policy.
And if you go on to assume that heartless future monetary policymakers will sabotage current attempts by the Fed to boost the future expected price level, then they would be even more likely to sabotage current fiscal policymakers, for whom that have a far greater distaste. So why didn’t Woodford get this result?
In the Woodford model once rates hit zero the monetary authority is literally speechless, except if they can signal changes in the future path of interest rates, i.e. promise to hold them at zero for an “extended period of time.” (Sound familiar?) But if they have a 2% inflation target (which is assumed) they have no incentive to do this after they have exited a liquidity trap. And the public understands this and hence doesn’t believe Fed promises to inflate.
Fiscal policy is different. They can do something meaningful right now. Even if short term rates stay at zero, fiscal expansion can boost AD in the normal Keynesian way (or by boosting velocity in the monetarist approach.) The fact that the inflation target stays at 2% in the long run doesn’t sabotage current fiscal policymakers, as current fiscal expansion it least is able to get you out of the deflation much sooner, and back on to that long run 2% inflation track.
So what’s wrong with this approach. Technically, there is nothing wrong with it. But it’s not what I consider an expectations trap. Monetary policymakers (especially Bernanke) know all this. They know that if they want to escape the liquidity trap quickly, they need to target the price level, not inflation. Bernanke said as much when he recommended that Japan do this. Woodford recently recommended a price level targeting approach. So at a minimum, a Fed that really wanted to be more expansionary would adopt a price level target, aka ‘level targeting’. They would promise to catch up for shortfalls. But if this is their promise, and they are assumed to renege on the promise, and sabotage current monetary policy, then fiscal policy is equally screwed.
If fiscal policy is to work, then it must raise AD, and hence the future expected price level. If the Fed won’t let them do that, then it won’t work. It doesn’t even matter if the short term rate is stuck at zero right now, and there is nothing the Fed can do right now to sabotage fiscal policy. Just the expectation that in the future they will act to prevent the price level from rising as the fiscal authorities hope, is enough to sabotage current fiscal policy. That’s why I started this entire overlong essay with the thought experiment about house prices, to try to convince you that what drives current assets prices, and current AD, is future expected monetary policy.
To summarize; future expected monetary policy is what drives AD. Fiscal policy may be effective, but only of the future Fed is expected to allow it to be. Current monetary policy may be effective at the zero rate, but only if the future Fed is expected to allow it to be. And there is almost no reason to expect that a future Fed (probably headed by Ben Bernanke) would try to sabotage and humiliate a current Fed that made a very public and explicit price level target in the midst of a severe crisis. Won’t happen. Period. End of story.
Tags:
8. June 2010 at 07:48
Fine Scott but ‘k’ arises from the time-path history of ‘r’, and ‘M’ is not the monetary base.
Therefore the solution is not so easy to acheive.
8. June 2010 at 08:13
Jon, Do you mean the “Cambridge k”?
I use the monetary base as my definition of the money supply. It’s more convenient as then you don’t have to worry about changes in the multiplier.
8. June 2010 at 08:47
I think in your last paragraph you mean “no reason to expect,” not “to reason to expect,” no?
8. June 2010 at 08:59
I feel like reputation or accumulated credibility can more than suffice to overcome the time inconsistency problem and I’m always surprised how often monetary economists today think time inconsistency is such a massive issue for monetary policy. Of course, theoretically it is a big issue, but in practice it seems that the issue is relatively easy to overcome if you’ve established that the central bank has been responsible in the past. I just find it strange that monetary economics spends so much time worried about the time inconsistency and all of that when it seems so easy to overcome in real life. Also, if we are going to talk about economics we must surely recognize the importance of incentives, and surely there is an incentive for central banks to not be vilified by the public down the road if they were to be renege on their promises in a few years. Thus, I don’t see how it’s a problem at all.
On Japan though, I think the tinkerbell theorists have a point now. While I agree this expectations trap wouldn’t have been a problem in the 1990s, I think it is in Japan now. I think the reputation of the BOJ is shot and I would question their credibility and commitment if they announced a 3-4% inflation target. It seems to me the BOJ doesn’t care about their reputation (doesn’t everyone think they are incompetent?), so you don’t have that mechanism to take care of the problem. I wouldn’t believe any commitment to a target coming from the BOJ, given their (apparent) historical aversion to even slight inflation. I think they would have to do Svensson’s exchange rate peg technique to credibly resolve the problem. Of course, this argument doesn’t apply to the United States where I believe our central bank still have enough believability / accumulated reputation and desire to be seen in a positive light that they could just commit to a 3% target and be done with it.
Now, you claim that the Woodford model works under dubious assumptions. I think that is wrong. For one, it’s very easy to modify those assumptions in just about any model that even resembles Woodford’s. But, more importantly, it appears that in practice those two assumptions have been entirely reasonable. Also, correct me if I’m wrong, but in the Woodford model doesn’t government deficit spending work because it acts like quasi-monetary policy anyway? By blowing up the budget deficit, households anticipate future inflation to pay of the debt. Isn’t government deficit spending just acting as a replacement of central bank committing itself to higher inflation?
Which then leads to another point. If you believe those two assumptions, fiscal policy should actually be worthless, at least the way I read it. If the Federal Reserve is not going to let inflation rise in response to a government deficit, then the fiscal policy chain is broken. Eggertsson actually made this point in a paper where he argued if the monetary and fiscal authorities are “goal independent” and the monetary authority doesn’t respond to fiscal pressure in any way, then the deficit spending multiplier is precisely zero. I don’t think there is any coordination between our Congress and the Federal Reserve, nor do I think the Federal Reserve bases policy around fiscal pressures – so shouldn’t the fiscal multiplier then be (at least approximately) zero? So, it seems to me if you make those two assumptions you actually can’t have any faith in fiscal policy, at least in the Woodford model.
So, I actually think your assumption about fiscal policy is too strong. You don’t even have to believe the Federal Reserve is going to counteract fiscal policy, you basically only have to believe it doesn’t care what the government debt is.
All and all though, I basically agree with you. The expectations trap seems more like a theoretical problem than a practical one, especially in the United States.
Oh, also, since you are an expert on the Great Depression, do you happen to know anything about Sweden’s experience with price level targeting during the Great Depression? I believe they were the first to escape the Great Depression, but I’m curious if you have any views on their price-level targeting during that time period since you brought it up right now.
8. June 2010 at 09:10
Scott: yes.
I happen to believe that understading what constitutes M is important. If people lose confidence in certain money substitutes what happens? What parameters change? Prices? Output? The market rate? Or perhaps nothing happens. If so were those money substitutes ever a part of M?
I agreed on tinkerbell, but I disagree that the quantity theory provides an adequate rational basis on which to form expectations.
8. June 2010 at 10:01
[…] Scott Sumner is making the same claim: If fiscal policy is to work, then it must raise AD, and hence the future expected […]
8. June 2010 at 10:41
Scott,
Again, it’s not obvious to me why the Fed needed to set a higher inflation target in the fall of 2008. 2% would have worked just fine if they had properly communicated thier commitment to hitting it. Core inflation was a bit above 2% in October of 2008. If they had just said “we’re going to pump enough money into the economy to ensure that core inflation averages close to 2% over the next three years”, I think that would have done wonders. But they didn’t do that. And core inflation in going to average closer to 1% or less from late 2008 to late 2011 according to the Fed’s own forecast. So they admit that monetary policy has been too tight, is too tight and will be continue to be too tight relative to their stated long-term target of 2%. The Fed chairman should have to take an oath “I solemnly swear to keep inflation at 2%, no more, no less, so help me God”.
The Blanchard idea of a 4% target so that equilibrium nominal rates will be higher is a serious turnoff for the world’s central bankers. All of this talk of higher inflation targets misses the point: what good is a higher target if you’re not committed to hitting it? Set a target, state that you arte completely committed to hitting the target and then demonstrate your commitment with action as needed.
Your thesis is that the Fed, the ECB and BoJ have been the largest contributors to this slump. Well what do they all have in common? They all lack a clearly defined symmetric inflation target. If the Fed and ECB both had hard targets of 2% and the BoJ had a target of 0%, the culpability of the central banks for the current state of the global economy would be obvious to many more people.
8. June 2010 at 14:22
@Gregor:
I’m not a fan of inflation targeting persay, but if we want believe that the growth trajectory of the Great Moderation was optimal (or at least worthwhile), then the Fed is the only institution with the power to increase NGDP expectations to a level consistent with that trend.
As of right now, the Fed’s actions have communicated that it doesn’t mind that we are below that trend, and IF they continue on with a 5% NGDP growth target (implying 2% inflation), then they are saying they are fine with setting a new path at a lower level. If that is optimal, then fine…but then the most fiscal policy can accomplish is damage control — the rest needs to be taken care of via grinding deflation (or if you prefer, recalculation).
8. June 2010 at 14:27
BTW: It may have been vain of me to assume you were talking about me in your post…quite obviously you were talking about Oliver Blanchard ;].
The way I understand it is that Oliver Blanchard wishes to have a permanently higher inflation target of 4% on the grounds that there is more cushion room for when an economic downturn happens — so that the Fed doesn’t run into the ZLB. The relative success of Australia would tend to validate this view, but I don’t agree.
8. June 2010 at 15:23
Scott-
I use the monetary base as my definition of the money supply. It’s more convenient as then you don’t have to worry about changes in the multiplier.
So you’re arguing that a permanent doubling of the monetary base would double the price level?
The price level is determined by the reservation at which people which to exchange dollars for goods. The price for automobiles is determined by the reservation price at which ordinary people wish to exchange dollars for cars.
If the Fed doubles the monetary base, what is the direct line of cause and effect that results in Joe Plumber being willing to spend $30,000 on a Honda Civic instead of $15,000?
Under certain circumstances, increasing the monetary base will result in more credit creation. Joe will have cheaper financing, and that will increase his reservation price. But credit creation is currently constrained by capital requirements and the lack of credit-worthy borrowers. Interest rates on automobile financing are about as low as can be. At the current time, banks are not reserved constrained in their ability to create credit (read the comments of this banker for example: “In recent months we have seen more deposits flow into our bank (the so-cslled ‘flight to quality’) and it has been hard to stay more than about 72-73% loaned up. ”
So how, specifically, does doubling the monetary base result in Joe Plumber doubling the price at which he is willing to buy a car?
8. June 2010 at 16:05
johnleemk, Thanks, I fixed it.
Ted, I mostly agree with you. I have argued many times that the fiscal multiplier is precisely zero if the Fed is doing its job, so any greater than zero estimate of the fiscal multiplier is an indicator of how incompetent the Fed seems in the eyes of the person making the greater than zero multiplier estimate.
I think Sweden did OK, although they really should have gone for reflation before stabilization. I recall they only pegged the price level for a few years. But the Swedes are certainly pioneers, they also pioneered negative rates on bank reserves, an idea I published early last year. And they were smart enough to stay out of the euro.
Jon, I wasn’t trying to defend the QTM, I was showing that even using the strongest monetarist assumptions, doubling the money supply may have no effect. I certainly agree that V and k are unstable at times.
Gregor, You said;
“Again, it’s not obvious to me why the Fed needed to set a higher inflation target in the fall of 2008. 2% would have worked just fine if they had properly communicated thier commitment to hitting it. Core inflation was a bit above 2% in October of 2008. If they had just said “we’re going to pump enough money into the economy to ensure that core inflation averages close to 2% over the next three years””
I agree. I meant that right now we need a greater than 2% rate to catch up to that trend line (we have fallen 1.4% behind.) I agree that 2% from Sept. 2008 would have prevented a severe recession (although there would have been a mild one.)
You said;
“Your thesis is that the Fed, the ECB and BoJ have been the largest contributors to this slump. Well what do they all have in common? They all lack a clearly defined symmetric inflation target. If the Fed and ECB both had hard targets of 2% and the BoJ had a target of 0%, the culpability of the central banks for the current state of the global economy would be obvious to many more people.”
Very good point.
Niklas, I do understand the argument that 4% helps you avoid a liquidity trap. It is a good argument. But Gregor also has a good argument. If the central banks are too conservative for 4%, just getting them back on the 2% trend line would help a lot. They’ve fallen well below it.
Devin, You’ve asked the question that lies at the very heart of monetary economics. It’s the fallacy of composition. Skip banks reserves for the moment and think about cash. If you double the amount of cash people have in their wallets, they want to get rid of some of it. But collectively they can’t, all they can do is pass it to someone else. Their attempt to get rid of excess cash balances drive up prices. When prices double they now want to hold the doubled supply of cash, and we are again in equilibrium. Adding bank reserves doesn’t really change things, unless you are in a liquidity trap—and those don’t last forever.
8. June 2010 at 16:13
Reasons to dislike stimulus spending: beside it doesn’t help end a downturn.
Giving the guy down the block a stimulus check now, with the promise to raise my taxes to pay for stimulus check next year, does little to really stimulate the economy. The guy down the block may increase spending with some small multiplier for the economy but I decrease my expenditures in the current period.
If you assume that people are interested in after tax income, those targeted will adjust how and when they receive income. Why take the same risks if the share you get to keep declines with a successful outcome? The more seriously they take the threats to their future income, the more they hit the brakes today.
Government may redistribute wealth but you are not growing the pie. And the returns on government investments are rarely superior to the returns of the private sector.
Perhaps the Europeans who are now tightening belts (or slowing the redistribution of wealth) have come to see that they need to follow a different path.
I wonder what Scott thinks that path will be, devalue the Euro? Expand the monetary base?
Simply, the government is spending a dollar to get 95 cents back. Times a couple of billion we begin to talk about real money.
Scott can argue, I think, that a surprise increase in the price level can make people feel wealthier (paying off their debts with now cheaper dollars).
I even suppose that like gamblers if people see a sudden increase in the price level, and they have assets and wages going up, they treat some of the increase like they are playing with house money. They suddenly have a bigger stack of chips in front of them and even if their relative wealth hasn’t increased, they act like it has.
I’m just afraid that the drop in the stock market, drop in housing value, combined with political changes – well you may need to put an awful big stack of chips in front of people before they want to gamble on the future again
8. June 2010 at 16:30
@Scott:
I agree that 2% would be better than the current policy (indeed, anything is), but even at 2%, we’re accepting a permanently lower path for GDP (assuming that the Fed will pursue it’s implied 5% NGDP growth rate going forward). My point is that with monetary policy committed to the new path, fiscal policy is there for the (perhaps valid) reason of cushioning the transition — but we shouldn’t delude ourselves into thinking that fiscal policy has the firepower to bump us back up to our previous trend.
This is where I believe Kling’s recalculation comes in, not so much as a cause, but as an effect of lower nominal spending.
8. June 2010 at 16:38
Scott-
I understand the quantity theory of money, and the impact of the quantity of money on the price level. What I disagree with is your specific claim. You claimed that doubling the money supply would result in doubling the price level. And then in the comments you defined the “money supply” as the monetary base.
The monetary base right now is ~$2 trillion. Household net worth is around $50 trillion. Total government backed paper (T-Bills plus insured deposits plus de-facto insured accounts, minus double counting) is probably on the order of $20 trillion. If you doubled the base, and simply mailed cash to all Americans, that would probably end up increasing net worth by around 10%. If you’re net worth is increased by 10%, are you going to double the reservation price at which you’re willing to buy a car? No way.
And of course, the effect on the price level depends on how you increase the base. If the Fed increases the Base by buying government securities, then private sector net wealth remains the exact same. If Joe plumber’s net wealth remains the exact same, is he going to be suddenly willing to buy a Honda Civic for $30K instead of $15K? No way.
So depending on how the Fed doubles the base, the net effect, in the current economic climate, would be a 0% – 10% increase in the price level.
So again, I just don’t understand what chain of cause and effect runs from doubling the monetary base to doubling the price level. Can you describe such a chain of cause and effect?
If you doubled the face value of all government backed paper, by simply marking every American’s accounts upwards (doubling everyone’s deposits, doubling every money market account, and doubling every T-Bill), then everyone’s net worth would increase quite substantially. They would then take that extra vacation, buy that car they were waiting for, and that would drive prices upwards. In fact, this is exactly the solution I would recommend for the current crisis.
8. June 2010 at 19:41
John Taylor on G20 and Stimulus
http://johnbtaylorsblog.blogspot.com/2010/06/is-g20-starting-to-get-back-on-track.html
Wonder what Krugman will say
8. June 2010 at 21:16
Scott:
“the Quantity Theory of Money””the idea that if you double the money supply, the price level will also double over time”
Yikes! Assume two countries both double their issue of money. One issues it for something of value (gold, land, bonds), and the other issues it by helicopter drop. The one that issued it for bonds (1) has exactly enough new assets to buy back the money it has just issued (2) only issues a dollar to people who are willing to part with a dollar’s worth of bonds (3) doesn’t affect the net worth of the people it bought the bonds from. We all agree that the helicopter drop causes inflation, but there are a few of us lonely souls out here who think that when a bank’s assets move in step with its issue of money, that money will hold its value.
8. June 2010 at 22:33
If I can join in on the Devin-Sumner discussion:
Scott:”Skip banks reserves for the moment and think about cash. If you double the amount of cash people have in their wallets, they want to get rid of some of it. But collectively they can’t, all they can do is pass it to someone else.”
Devin:”So again, I just don’t understand what chain of cause and effect runs from doubling the monetary base to doubling the price level. Can you describe such a chain of cause and effect?”
Scott has described one chain of cause and effect, but this is not at all what’s happening. What Scott describes is a helicopter drop where money base is given to household. The massive injections of money base were done by buying MBS and other assets for reserves. There’s no mechanism where households don’t want to hold money and have to pass on their excess reserves by spending cash. The situation is almost completely the opposite of Scott’s- Banks are holding massive amounts of excess reserves and are NOT passing them along. These excess reserves have persisted and moved almost in lock-step with changes in the money base. So this mechanism has no bearing on the current situation. Naturally there will be a time when banks will lend out their excess reserves- that day is not today and there’s no indicator we’re moving in to that day anytime soon. Until then, this charming story bout households is just a fairy-tale.
In one sentence- this is not kosker: “Skip banks reserves for the moment and think about cash.” This has nothing to say about the current situation.
Hick’s view of the liquidity trap was that bonds and cash were perfect substitutes. Look at money base and excess reserves, and Hick’s liquidity trap looks quite reasonable:
http://research.stlouisfed.org/fred2/series/EXCRESNS
http://research.stlouisfed.org/fred2/series/BOGUMBNS?cid=124
I’m sure Scott will continue to lean hard on a quarter of a percent interest rate on reserves. It’s quite a small number to explain over a trillion dollars in excess reserves. (Notice that the entire money base was only about 800 billion pre QE) The base has more than doubled. You can decide- liquidity trap or a quarter of a percent.
9. June 2010 at 01:59
As adults do not believe in fairies, Tinkerbell won’t get us to fly. It is well known that Presidents, Central Bankers and High Financiers believe in Chauncey Gardiner.
“…suddenly the President addressed him: ‘And you, Mr Gardiner? What do you think about the bad season on The Street?'”
How Chance replied is unimportant. How the President and Mr Rand reacted is.
“‘I must admit, Mr Gardiner,’ the President said, ‘that what you’ve said is one of the most refreshing and optimistic statements I’ve heard in a very, very long time.'” (Being There pp52-53).
We need a frame changer.
9. June 2010 at 03:48
DanC, I agree that fiscal stimulus doesn’t do much. As far as monetary stimulus, I don’t think inflation works by making people feel richer, I think it works by lowering real wages.
Niklas, I just don’t see how fiscal policy can do much if the Fed is committed to a contractionary path. Fiscal policy only works if it raises inflation. What if the Fed won’t allow that?
Devin; You said;
“The monetary base right now is ~$2 trillion. Household net worth is around $50 trillion. Total government backed paper (T-Bills plus insured deposits plus de-facto insured accounts, minus double counting) is probably on the order of $20 trillion. If you doubled the base, and simply mailed cash to all Americans, that would probably end up increasing net worth by around 10%. If you’re net worth is increased by 10%, are you going to double the reservation price at which you’re willing to buy a car? No way.”
I’m glad you mentioned this because this is not at all what I was arguing. You are describing fiscal stimulus, not monetary stimulus. I envision swapping the cash for other financial assets. This would not increase wealth at all. Monetary policy doesn’t work because it increases wealth, it works by lowered the real value of the asset that increases in quantity. If you double the amount of apples the value of each apple will fall. It has nothing to do with whether apples are a big percentage of GDP or a tiny percentage. If you double the amount of cash, the value should fall in half. That’s basic monetary economics. And it has absolutely nothing to do with making people feel richer.
You asked:
“So again, I just don’t understand what chain of cause and effect runs from doubling the monetary base to doubling the price level. Can you describe such a chain of cause and effect?”
Again, this is where the fallacy of composition comes in. If you assume that each person demand for base money is independent of Fed policy (and mine sure as hell is) then if the Fed doubles the money supply I won’t want to hold more real balances. I’ll try to get rid of excess cash balances. But society as a whole can only do that by bidding up prices. It’s the same mechanism that makes apples worth less when there is a doubling of the supply of apples.
BTW, This is not my theory, it is just standard monetary economics (monetarist and new Keynesian.)
I agree that your last idea would work, indeed even if just 10% higher. But try convincing unions that their workers real pay should be cut in half.
Thanks for the Taylor link DanC.
Mike Sproul, Yes, I recall you don’t buy the QTM. I certainly agree that OMOs don’t affect wealth in any direct way. I use an excess cash balances mechanism. I’d rather go shopping with cash than T-bills in my wallet. So if the Fed stuffs my wallet with cash, and banks don’t want to hold ERs because they can get a positive yield on other assets, then prices have to give. (Of course that mechanism is broke right now.)
Just another economist, I’m afraid you forget of read the last line of my first response to Devin. I conceded that the analysis I was giving did not apply to liquidity traps.
Regarding excess reserves, I have not argued that a quarter point would necessarily make much difference. I have argued that 2 1/4 points would make a big difference. If banks could earn 2 1/4% more interest on T-bills than ERs, they’d rapidly lose interest in holding ERs.
One final point. If you assume the liquidity trap lasts forever, then I agree that monetary theory would have to be redone (including all perfect substitutes for base money.) But the odds of the liquidity trap lasting forever seem very remote to me. I assume the Fed also feels this way, as they feel they must keep reassuring markets that they plan to tie up all those ERs before they start creating inflation.
W le B, Yes, it is not enough to say you will create future inflation, you must do it.
Everyone, My argument doesn’t even require the QTM is true. I just set it up as an assumption so that I could show that temporary currency injections have no effect. They still have no effect if the QTM is not true, so nothing in my post hinges on that. I was just trying to explain to people who do believe in the QTM why we aren’t getting inflation right now.
9. June 2010 at 05:33
Scott,
(and Niklas, I think this address your point as well),
You said:
“I agree. I meant that right now we need a greater than 2% rate to catch up to that trend line (we have fallen 1.4% behind.) I agree that 2% from Sept. 2008 would have prevented a severe recession (although there would have been a mild one.)”
That’s true. As you often point out, Bernanke has written many papers or the importance of the “price level gap”. But politically, the Fed announcing a period of above 2% inflation followed by a restoring inflation to 2% is not going to happen. And right now the Fed’s own forecast is for below 2% inflation for as far as the eye can see (which, when you really think about it, is a bit of an outrage). What I’m saying is that announcing a commitment to returning core inflation to 2% quickly, say, within a year, (and keeping it there), would be a much more simulative policy than what we have right now and might be enough to ensure a healthy recovery. It also has the advantage of being simple. No one can accuse the Fed of being overly dovish if all they’re trying to do is hit their 2% inflation target. It might not be the optimal policy, but I think it would be a huge improvement.
One advantage of a single-mandate inflation target is that it makes monetary errors symmetric and obvious to everyone. If core inflation is 0.6% at the end of 2011 and at the end of 2012 (the low end of the Fed’s forecast range) many people would argue that the Fed is still fulfilling its commitment to “price stability”. In Canada, the BoC Governor would have been fired long before 2012. The ECB has an even worse mandate with its asymmetric “below 2%”. If we went through another Great Depression, the ECB would still be fulfilling its mandate. The mysterious BoJ is obliviously worse still.
Yes Niklas, I was referring to Oliver. And I agree with you, I don’t think the facts fit his thesis. The ECB still has “conventional” policy room left but refuses to use it. And whenever major central banks announced asset purchases, the announcements were greeted very favorably by equity and commodity markets. So central banks can still stimulate at the ZLB and just because they’re not at the ZLB it’s no guarantee that they will get policy right. A target is only as good as a central bank’s commitment to hitting it. But the more clearly defined the target, the more obvious it is when they’re not doing their job.
9. June 2010 at 08:59
Thanks for the excellent reply.
I didn’t notice the last line of your response, you are correct. However, we are in a liquidity trap by any measure.
Let’s take a look at an excellent paper on the liquidity trap: Sumner 2002:
“A country experiencing extremely rapid money
supply growth, weak aggregate demand, and extremely low nominal interest rates would seem to best fit the description of a liquidity trap. I use the term “extremely rapid” because fiat money creation is essentially costless. A country “stuck” in a liquidity trap is a country
where the central bank has tried hard to boost aggregate demand and has failed. Given the ease of money creation in an unconstrained fiat money regime, growth rates of a few percentage points are hardly sufficient to demonstrate serious effort. The BOJ certainly has no lack of government debt to purchase.” (p. 487)
The current situation in the United States passed what I’ll call the “Sumner test” 🙂 with flying colors. We have ” extremely rapid money supply growth, weak aggregate demand, and extremely low nominal interest rates” Has your position changed since this article, or do you disagree that the United States fits the “Sumner test.” I don’t think revising your position is contradictory- it’s a sign of flexibility, openness, and honesty, butit should be defended.
Also, I think this needs saying. It’s a liquidity trap, not a liquidity black hole. You can get out of a trap- it’s temporary, just like you said. There are several ways, including an exchange rate anchor a la Svensson. This would have been a good way for Japan to escape their liquidity trap as they can easily increase exports to boost aggregate demand. Also, this is consistent with the FDR devaluation and recovery. Another way is to use fiscal policy, as long as there is plenty of monetary base available to support the resulting nominal GDP. This would be preferable today, as the US’s size makes it so that exchange rate changes have a smaller effect.
So there is no need to buy up all the debt, etc. as this will result in a hyperinflation once the liquidity trap is exited. Central banks are currently too concerned with inflation- but that said, once velocity returns to normal levels, I wouldn’t like for the monetary base to remain permanently doubled. It is a fine line to walk, as central bankers should sign higher inflation to reduce real interest rates, but keeping all those excess reserves in the system will spark uncomfortably high inflation. So far, they are too contractionary.
9. June 2010 at 09:14
@Scott:
“Niklas, I just don’t see how fiscal policy can do much if the Fed is committed to a contractionary path. Fiscal policy only works if it raises inflation. What if the Fed won’t allow that?”
That is exactly what I’m saying. If the monetary authority “deems” it acceptable to stick with a lower level path for NGDP, then all fiscal policy can hope to do is “round out” a few jagged edges in the transition by cushioning the hardships (unemployment) that arise from disequilibrium in the face of sticky prices.
I think DeLong has a good point about borrowing costs, but that in-and-of itself doesn’t make the case for the government borrowing, because fiscal policy can’t be aimed at boosting inflation if the monetary authority is targeting inflation (or NGDP) from a lower level. However, the case can probably be made that we as “a society” should invest in some things that we “need” while borrowing costs are abnormally low. Of course, that “need” is subjective.
So, if during a recession we let the target path of NGDP fall by $1.3tn, and just kind of pick up our boots and trudge the same rate of growth from the new level, fiscal policy isn’t going to magically boost us back up to the old level.
10. June 2010 at 03:22
@ Just Another Economist, but is there “extremely rapid money supply growth”? I think this blog recently pointed us towards http://www.telegraph.co.uk/finance/economics/7769126/US-money-supply-plunges-at-1930s-pace-as-Obama-eyes-fresh-stimulus.html
10. June 2010 at 04:13
Guys, I hate to barge into your discussion about what is going on in monetary policy and what is expected to happen, but you all seem to consistently ignore the role of expectations in relation to the fiscal deficits. If there is no action, or even sign of action, on fiscal deficits consumers will be reluctant to spend and businesses be reluctant to invest. It will be prudent to conserve cash and/or look for inflation-proof but low risk assets until the big issue of structural government deficits is resolved. Resolution will come eventually, either deflation and restructuring (the UK throughout the nineteenth century), or inflation and restructuring (the “buy time” approach to slowdowns for the last twenty years now no longer viable), or (disaster of disasters) inflation and no restructuring (stagflation / Zimbabwe / UK in the 1970s).
Monetary policy may be a necessary condition to enable a recovery, but confidence in one’s government to live within its means is arguably more necessary. Nothing I see or hear from the US seems to point to any sort of fiscal consolidation for at least 12 months, this is unlike the situation in most of Europe where it is very much on the agenda, even if there are not many achievements to point to just yet – except perhaps in Ireland.
Your discussions all seem to be beside the point until Washington gets enough pressure from the voters on the fiscal deficit.
10. June 2010 at 06:06
> Unfortunately my brain is not wired properly to understand monetary policy based on manipulating real interest rates.
Unfortunately their brains are not trained to understand it either.
This is what engineers call a control system problem. Do any economists actually have training in control system theory? No. Not a one. As an engineer I can tell you that even the most trivial control system can have all sorts of strange and paradoxical relationships and behaviors, that is they seem that way if you don’t know control system theory and the math behind it. Do X to an input one time and getting Y, but doing X another time with a subtle system parameter change and you get Z or -Y or 10Y or something else unexpected, IS EXPECTED AND COMMON BEHAVIOR FOR SUCH SYSTEMS.
Most people think linearly: X gives Y so 10X gives 10Y. The world is NOT linear but can be approximated as linear in very small differences. Fooling yourself into thinking linearity lives forever is as naive and ignorant as imagining exponentials can as well.
Economic and monetary systems are very clearly far more complex versions of control systems. They are nonlinear multivariate differential equation systems. Engineers typically have the luxury of controlling things by simply deciding: the system will NOT be nonlinear and multivariate. The design strategy of “don’t do that” works pretty well. There are times when, like with economic systems, you don’t get a choice like that. From those experiences we’ve found (with the ample help of mathematicians and physicists) that such system can be controlled but less so, more dangerously and with far, far less certainty of control.
One look at the models typically used for both fiscal and monetary policy and it’s shocking obvious that these models 1) are very naive, 2) so much so they should probably not ever be used for anything in real life, and 3) using them at the wrong time or situation, specifically in times of transition and non-equilibrium, is at best foolhardy and at worst completely criminal.
The level of mathematical ignorance in the economic profession is quite stunning – on par with allowing kindergarteners to perform space shuttle maintenance and then expecting things to end well!
10. June 2010 at 11:07
Gregor Bush, Yes, those are exactly my thoughts. Well said. I am as perplexed and frustrated as anyone.
Just Another economist. You said;
“The current situation in the United States passed what I’ll call the “Sumner test” with flying colors. We have “ extremely rapid money supply growth, weak aggregate demand, and extremely low nominal interest rates” Has your position changed since this article, or do you disagree that the United States fits the “Sumner test.” I don’t think revising your position is contradictory- it’s a sign of flexibility, openness, and honesty, butit should be defended.”
I do think the US could be considered a liquidity trap by some definitions, and I would say ther same about Japan. That paper was probably written in 2001, before the big increase in the Japanese monetary base. Of course in the case of the US, I never anticipated the decision to pay interest on money. At the time I defined base money as being non-interest bearing, and so by that definition there has been no big increase in the US monetary base.
But I would also point out that even if a country is in a liquidity trap, There are ways out other than printing money. These include currency depreciation and/or CPI or NGDP targeting, especially futures targeting. If the Japanese wanted inflation, they would have done one or the other. But the BOJ obviously does not want any inflation, so the liquidity trap (if you wish to call it that)is “voluntary.” But I agree that in the US today, or in Japan in 2002, a substantial increase in the monetary base might have had little effect.
I think the paragraph you cite might have had a slightly different meaning from what you thought. In that paper I argued that a central bank could always inflate under a fiat money regime, if they were determined enough. I also argued that a necessary (but not sufficient) condition for someone to claim the BOJ was trapped was that they had at least tried to escape. As of 2001 when I wrote that paper the BOJ had not yet even tried QE. They did try and although some claim it succeeded, I think that is debatable. It did succeed in stabilizing the CPI for 5 years, but the GDP deflator continued to fall.
In 1993 I argued that temporary currency injections would not create inflation. The 2002 currency injections in Japan were temporary, so it’s no surprise they didn’t cause inflation. I think the US currency injections are also likely to be temporary, and that (plus IOR) explains why they haven’t created inflation. In general, I don’t think QE is the best way out of the liquidity trap. In that paper I used the reductio ad absurdum of truly massive QE, but no real world central bank would do that. It would work, but will never be tried.
I don’t agree that fiscal stimulus is a good way out of a liquidity trap. Yes, it might work, but it is very costly. Price level or NGDP targeting are far superior.
I agree with your final point that permanently doubling the base would lead to high inflation, and would be a horrible idea. Indeed if it occurred, the high inflation would happen almost immediately, not with a long lag.
Niklas, I agree. DeLong’s idea reminds me of an earlier post I did where I argued that what Krugman calls “depression economics” is actually “expected depression economics.” It’s not the economics for when you are in a depression, but rather for when you think you will be in a depression in a year or two. But with a well run monetary policy you should never be expected to be in a depression in a year or two, and hence you should never use fiscal policy to influence AD or to pay for projects that wouldn’t meet a cost/benefit analysis in normal times.
W le B, I think he meant the monetary base, which is how I define money. But to be honest I don’t recall what definition I used in that article. Neither the base nor the broader aggregates had grown explosively in Japan by 2001, unless I am mistaken (which I may be.)
James, I agree that fiscal deficits are worrisome, but I argue that monetary stimulus is the best way to reduce fiscal deficits in the short run. In the long run we need to reform entitlements. That’s for the US, I know less about the UK, but would guess that the need to reduce deficits is at least as great, if not greater, than in the US.
JSG, You said;
“The level of mathematical ignorance in the economic profession is quite stunning – on par with allowing kindergarteners to perform space shuttle maintenance and then expecting things to end well!”
It’s pretty obvious that you don’t know much about non-linear models in economics. But just because you don’t know about them, doesn’t mean they don’t exist.
I don’t think I assume linearity when I make recommendations for monetary policy. If you see me assuming linearity someplace, please point it out. I certainly don’t think all economic relationships are linear, nor have I ever met an economist who dies. I don’t regard the assumption of linearity as a big problem in economics, there are much more serious problems out there.
10. June 2010 at 12:51
When Nick Rowe says, “Maybe monetary policy is about the supply and demand for money?” is that a serious question? Are there really people who don’t think monetary policy is about the supply and demand for money? What the heck else could it be about?
Scott, I wonder about James in London’s point. If the public gains confidence because the politicians finally address the deficits in a serious way, might that not reduce the demand for money and result in a de facto easing of monetary policy? Along the same lines, the business owners I speak with are reluctant to make any long term commitments because of policy uncertainty. If we removed that uncertainty somehow, wouldn’t that be the equivalent of an easing of monetary policy?
10. June 2010 at 16:00
[…] Source […]
10. June 2010 at 19:34
The willing suspension of disbelief is the theme of this line of reasoning. Take a bow tinkerbell. If you think you can fly, you can fly. Nothing wrong with that if it works and it usually does, except when it doesn’t. Economics is the land of Oz to mix a metaphor. Therefore, reasoning in the normal sense just doesn’t work so you end up with a logical construct that leads to opposite conclusions. Money is like that because it is not logical. Who in his right mind would exchange a cow for a piece of paper? I might if I thought that someone else would take the piece of paper in exchange for something I wanted. Do that enough times and you begin to believe that the paper has real value. You then start to make long involved arguments about monetary policy that seem to make sense, except when they don’t. That’s why economists are not rich. They literally do not know what they are talking about. That is they do not understand all the assumptions express or implied in their arguments.
10. June 2010 at 19:40
Why the argument about this? Money has no intrinsic value even when gold is used for money. It only has value because people believe it has value. You are talking about something like the boiling point of water that is constant. You are talking about human psychology and the very act of talking about it changes the way people think about it. The way they think about it changes the way they behave. If they are worried about having money coming in, then if they are rational, they save. Of course, they may not be rational and decide that if they are not going to have money coming in they might as well spend everything they have. If people believe that home prices were too high because the prices were not supported by wages they had, then they should not invest in houses expecting a short term profit unless they believe that wages will increase or that they can borrow money regardless of what their wages are. Anyway in economics Tinkerbell is right if you think you can fly then you can. (But don’t try this by jumping off the garage.) Analogies can be can be carried too far.
11. June 2010 at 05:07
Joe, You asked;
“When Nick Rowe says, “Maybe monetary policy is about the supply and demand for money?” is that a serious question? Are there really people who don’t think monetary policy is about the supply and demand for money? What the heck else could it be about?”
I’d say about 99% of people think it’s about the supply and demand for credit, and the price of credit (i.e. the interest rate.) Only a tiny fraction of us believe it is about the supply and demand for money, and the price of money (which is one over the price of goods.)
Yes, lower demand for money or any increase in confidence can indirectly make monetary policy more expansionary. Of course if the Fed is doing its job those things shouldn’t matter. But if it isn’t, then stronger expectations can help policy.
Ben Goff, There is no mystery to money having value, it is a useful medium of exchange. It has value for the same reason as wallets have value–they make shopping more convenient.
11. June 2010 at 05:59
[…] indeed, okay with allowing the economy to downshift to a lower trend growth path. As I said in a comment to a recent post by Scott Sumner: If the monetary authority “deems” it acceptable to stick with […]