In praise of backseat drivers

Here I’ll try a couple analogies to better explain earlier posts on causation and NGDP targeting.  As you will soon see, my literary skills are at the 5th grade level.

Captain Ben is piloting a large ocean-going cruise ship toward the open ocean, but must steer between two sandy headlands.  A steady wind is pushing the ship to the left, toward one of the sand-bars.  The captain has plenty of time to adjust the rudder and engine controls to counteract the wind, but reacts too slowly and the ship continues to drift to the left, eventually getting stuck on the sandbar.  Did the captain cause the accident through negligence, or did the wind cause the accident?

Not much of an analogy so far, but let’s add a few new elements to correspond to my earlier causation post.

Now assume the control room is full of a couple dozen other experienced ship captains who are on the cruise, and one trainee.  None of the ship captains suggest Ben is doing too little to offset the wind; most say he is doing all he can and a few even suggest he is doing too much, arguing “we don’t really know whether this wind will persist.”  Only the trainee thinks the Captain is doing too little, and he becomes increasingly agitated as the ship seems to be drifting to the sandbar on the left.  His suggestions grow more and more frantic:

“Turn the wheel further to the right; you haven’t turned it as far as it will go.”

“Turn on the auxiliary quickfiring ejectors (side mounted water jets that help steer the ship in close quarters.)”

“Don’t you see we’re headed right for the sandbar!”

Eventually the captain did turn the wheel all the way to the right, but the trainee kept on babbling on about the quickfiring ejectors:

“If you wait till we hit the sandbar the QEs won’t be nearly as effective!”

“You must act now!”

The other ship captains mocked the trainee mercilessly:

“That’s right, give her all she’s got, Scotty”

Some whispered that he had gone to one of those fresh water training schools, where they learned how to steer “boats” on the Great Lakes.

“That poor fool just doesn’t understand what’s involved in steering the big ships in salt water.”

“Yeah, he’s originally from the Midwest.”

Eventually the ship did get stuck on the sandbar.  The passengers were told it might be months before the ship could be dislodged.  There was great concern about what would happen to their jobs after such a long period of unemployment.

Now the QEs were finally turned on, but the ship responded very slowly, as most of the hull was still stuck on the sandbar.

“Ha, just as I thought, those QEs are not very effective; I knew that trainee was babbling nonsense.”

OK, the story is both silly and self-indulgent, but what about the question of causality, did the captain cause the accident?  In my earlier post (here) I argued that it was hard to assign moral culpability if most other experts would have done the same.  So Captain Ben cannot be “blamed” for the accident.  On the other hand if there is an alternative steering “policy” that would have prevented the accident, and if the captain had plenty of time to react, then one could say the accident was caused by not setting the steering mechanism at the right position.  I don’t think anyone would be so foolish as to argue the captain could only be said to have caused an accident if he turned the steering the wrong way.  Even a policy of holding the steering wheel steady could be said to cause an accident when there is an obvious bend in the river straight ahead.

So how do we prevent these accidents in the future?  The answer is to encourage “back seat drivers.”  Yes, I understand that my overwhelmingly male readership is almost all composed of above average drivers, make that well above average.  And I’m pretty sure most of you don’t like back seat drivers.  But is there really anything wrong with an occasional nudge?

I’m not going to spell out the NGDP targeting plan again, it’s discussed here.  But I did notice that many commenters were concerned about what would happen if nobody entered this policy market.  My response is that nothing bad would happen.  At the beginning of each day the Fed should set the base at a level expected to produce on target growth (say 5% NGDP growth.)  If no one enters the market, then we can presume that traders think the Fed is doing a pretty good job.  And that would be fine.

On the other hand last September and October it was obvious to me that NGDP growth was going to come in well below target.  I might have waited a few days to see if the market jumped in, but if they didn’t I would have grabbed the wheel and yanked it to the right with all my strength, or should I say all my dollars.  We’ve all complained about Fed policy, wouldn’t it be nice to be able to put our money where our mouth is.  Why should just one captain, or even twelve, have all the fun.

Yes, I know that 99.9% of the time backseat drivers are annoying.  But there is always that one time when they point out the child, or dog, or red light you didn’t see.  In that case isn’t it better that they say something, rather than keep their mouth shut?

Tags:

27 Responses to “In praise of backseat drivers”

1. colin
24. May 2009 at 20:07

so are you going to tell us how to differentiate between the quality of back seat drivers? we didn’t listen to Roubini or Krugman and we’re the worse off for it. but how should we have known to pick out those two among the presumably many actual back seat drivers? and they were at least renowned prior to the crisis. you are even more obscure. it doesn’t look like we’re set up well as a society to guide this ship. Cowen and Hanson would probably say something about signaling right about now…

25. May 2009 at 00:48

Scott,

you might be interested in this from Cochrane:

http://faculty.chicagobooth.edu/john.cochrane/research/Papers/fiscal_crisis_princeton.pdf

and slides:

http://faculty.chicagobooth.edu/john.cochrane/research/Papers/fiscal_crisis_slides.pdf

The last slide is titled “Intellectual Casualties” and lists MV=PY as the first.

3. Bill Woolsey
25. May 2009 at 02:53

Great story, Scott.

Suppose instead that it isn’t a wind, but rather a current that is pushing the ship against the sandbar.

The current is caused by a dam that has failed up river on a side channel. Rather than steer the boat against the current and continue out to sea, the captain is looking up river at the dam (water is spilling out rapidly.) He is on the radio giving instructions to the construction workers trying to plug the leak in the dam.

The young trainee is shouting, turn, turn. The captain and all of this buddy captains are looking to the side, up the channel at the dam. They are all discussing how to fix the dam. Why the dam failed. What could be done to strengthen the dam to avoid future failures.

They all see the rush of water and the current they are in. (From their perspective, it is pushing them back, looking over to the side like they are.)

Turn, turn, says the trainee. Turn on the QE.

Shut up kid. Once the dam is fixed, the current will stop, and we will head out of the harbor without any trouble. Hey gang, maybe we could put one of the QE jets on one of our landing boats, and send it up the channel. The construction workers maybe could use it….

And yes, turn the rudder a couple of points…

Hey, our landing boat is having trouble getting up stream to the dam. Let’s turn on our QE jets and push it.

What? What? says the trainee. That is going to push us even further towards the sandbar. Are you crazy? Turn.. Turn..

Shut up kid. Once the dam is fixed, it will be OK. Um, turn the rudder a half point….

Scott:

The Fed was trying to fix the shadow banking system. If the shadow banking system was fixed, then it is likely that the demand for base money would be lower. They paid interest on reserves to obtain funds to fix the shadow banking system. I could have worked….

I left out of our imaginary conversation the relationship between the dam owners and owner of the ship. How important the dam is to the whole community. Not only does it provide lighting to our homes, but power is exported to the foreigners. (The shadow banking system played a key role in credit markets and America’s comparative advantage is supposedly in high finance centered on Wall Street.) The way the dam owners traditionally give huge birthday presants to the owner of the ship. (Crazy, isn’t it? Oh, all those campaign contributions to both parties.) Yes, maybe the captain could save the passengers trouble and get them out to sea, but what about the birthday presents going to the ship owners? And what will we do without the dam?

How about rather thqan passengers going out to sea, the ship is full of fuel and coming into port. Failure to bring the ship into port will result in the construction workers on the dam being unable use power machinery. Further, the generators that people will need when the dam no longer provides power, will have no fuel.

No dam failure, no problem. Focusing on the dam failure, the ship captain failes to steer the ship to port. Failure to get the ship to port not only makes the problems created by the dam failure worse, it also causes the dam to fail worse.

4. Bill Woolsey
25. May 2009 at 07:36

I must admit that I didn’t finish Cochrane’s paper yet. However, he includes MV = Py in that paper. He puts that in to capture the special chacteristics of money.

He claims that this equation, along with the one that represents the government fical equilbirium condition must agree on an inflation rate.

One possiblity he mentions (the relevant one) is that M is determined to keep P on target. And then taxes are adjusted to pay off interest and principle on the national debt. (He calls this a Ricardian policy!!!)

In my view, Cochrane epitomizes the failure of monetary economics based upon finance principles (money as an investment) and intertemporal optimization.

5. Bill Woolsey
25. May 2009 at 08:00

I finished the paper.

MV = Py _with constant velocity_ is finished!!!!!!!!!!

Who believed that?

In 1956, Yeager wrote about a scenario in which an excess demand for government bonds pushed interest rates on government bonds so low, that the excess demand for bonds at that point shifted to an excess demand for money. Under those circumstances, open market operations (purchasing government bonds with money) exacerbate the excess demand for government bonds, shunting even more of it to money. The increase in the quantity of money will be matched by an increase in demand.

I don’t really see what the “fiscal” approach has added to this.

Frankly, it was difficult for me to take Cochrane too seriously after he said that the fiscal equation determines the price level even with M=0.

Cochrane’s research program fails out of excessive formalism. More generally, from the notion that money can be treated like shares of stock in the govenrment.

As he seems to understand sometimes–money is especial.

It seems to me unlikely that the perfect substitute relationship between money and bonds would hold true as the quantity of currency falls to zero.

This model ignores the possiblity of explicit default on debt. Also, the focus on expected future tax increases appeares to ignore the possiblity of cuts in other sorts of spending.

Finally, the “depression” connection with inflationary expectations involves efforts to defend the exchange rate.. or so I think. There was nothing in Cochranes model that suggests that a drop in velocity will not raise aggregate demand.

25. May 2009 at 08:07

Bill,

doesn’t section 3 argue that MV = PY can’t make sense of the current situation with or without constant V?

Here’s a quote from page 5 in section 3:

“In short, financial institutions didn’t want more money and less bonds. They wanted more of both, and less of other assets, and in massive quantities. The “special” or “liquidity” services we usually associate with money applied with nearly equal force to all government debt to these actors.

MV(·) = PY does not really allow us to address this sort of event.”

25. May 2009 at 08:11

Of course, Bill, if you want to dismiss Cochrane’s whole research program that’s your perogative. But keep in mind that your also dismissing agruments from Chris Sims, Mike Woodford and I guess some other.

These are some awfully smart guys for you to be so sure they’re all wrong.

8. 123
25. May 2009 at 10:33

The problem is not the permanent lack of interest in NGDP market. The problem is that under your proposal the wheel is tied to the NGDP forecast in a strange and impractical way. Volumes of NGDP market are correlated not only to the forecast itself, but also to the changes in the bond market conditions. It is a sad fact that even this imperfect scheme would have avoided the worst of captain Ben’s crisis.

By the way current capital markets provide plenty of opportunities to profit from Fed’s mistakes.

9. Thruth
25. May 2009 at 11:35

123: I still don’t get the point you are making. How would YOU design the NGDP policy/market?

10. Lord
25. May 2009 at 15:32

I am not sure most of those backseat drivers were not saying as so many continue to say even now ‘whatever you do, don’t turn further right, it’s the bar on the right we have to worry about, the one on the right’.

11. ssumner
25. May 2009 at 17:24

Colin, I think you are new here. Check out my NGDP futures market post, which explains the idea. The link is in this post. Basically anyone with money can play the game.

AdamP, Cochrane said MV=PY with constant velocity and long and variable lags is a finished. Hooray!! That pretty much summarizes what I have been trying to do with this blog for 3 months. Finish off that model for good. The right needs an effective counterweight to Krugman, and with all due respect I don’t think the monetarists are providing one.

I also strongly agree with Cochrane’s opposition to fiscal inflation. He’s right that it will result in higher taxes and reduce growth. That’s why I oppose fiscal inflation and support monetary inflation. My only complaint is that he limits his analysis of monetary policy to conventional options. There are lots of non-conventional monetary stimuli that are highly effective at zero rates and reduce the budget deficit. He’s right that we shouldn’t be raising the budget deficit.

Regarding Bill’s criticism of the fiscal view, I basically agree. It seems to me that the fiscal view is only relevant for banana republics that face the threat of hyperinflation. Despite all the sarcastic remarks out there recently, the US is not at that point yet. But if we continue with contractionary monetary policies I suppose anything is possible.

Bill, Those are very good additions. It occurred to me that if I spent more time there are other factors I could have added. But the trickiest is this—the “wind speed” depends on expectations of the captain’s steering, and I don’t think there are any nautical analogues for that.

123, You keep saying “volumes” are affected, and I still don’t see why that would greatly distort the risk premium. Slightly maybe, but if it moved too far it would bring a lot more speculators in.

I don’t understand your point about profiting from Fed mistakes. If (today) I expect NGDP growth to be less than 5%, how do I profit? Everyone knows it will be less than 5%–it’s priced into assets.

Lord, I pay no attention to all those pundits that say inflation is coming. Money speaks louder that words–why aren’t they buying TIPS? The same people who are in the TIPs markets, and all the other markets with bearish sentiments, would be in the NGDP futures markets.

25. May 2009 at 20:12

Scott,

Cochrane clearly says that MV() = PY is not useful in the current situation with ANY velocity specification. He doesn’t just criticize the constant velocity version.

13. Nick Rowe
26. May 2009 at 01:31

I just read (OK, skimmed) the first Cochrane paper. A very interesting read. In my view, full of great insights, and also horrible mistakes.

It would be very easy just to trash him for those mistakes. But also a mistake to do so, I think.

As an example of a mistake, his equation 1 implies that a country with a fixed stock of money and no government surpluses ever must have an infinite price level!

As an example of an insight, his recognition that expected future fiscal policy must impact AD today.

14. Bill Woolsey
26. May 2009 at 03:06

Nick:

“future fiscal policy must impact aggregate demand today.”

That is mistaken. Most importantly, it depends on monetary institutions. With some monetary institutions, future fiscal policy is irrelevant to aggregate demand today.

His particular model requires the assumption that government debt will be monetized, impacting the future price level, thus impacting todays money demand, and given the quantity of money today, will impact aggregate demand today. Lots of problematic steps in there.

What is the price level when M = 0?

Cochrane appares to realize that it is constant V that is the problem. The existing situation is that there is
was an increase in the demand for money and the demand for treasury bills. Velocity dropped. The problem is that an increase in the money supply implemented by open market purchases of the govenrnment bonds in excess demand at zero interest rates will result in the velocity falling to match the increase in the quantity of money.

See.. I did it using MV = Py.

Now, Cochrane has a model where M and B show up together (M+B) I suppose that does apply nicely to the special case were there is an excess demand for both. It also applies (and was created) for the situtation where M will be used to pay off B. It is a model of inflationary default. (Explicit default is ignored.) OK… two special situations where it might apply to something. Oh, by claiming that the discount rate of expected goverment surpluses changes, he can create deflation.

OK. The interest rate on “government bonds” is not zero. It is close to zero for T-bills. So, “money” isn’t a perfect substitute for “bonds,” only T-bills.

The problem of the Fed focusing on the interbank lending rate and implementing this through trades in T-bills _is_
the problem. By combing all bonds together and assuming that what drives the price level is the probability of
deflationary default is jsut a red herring.

By the way, I think Krugman is smart too. I am sure there are physicists who are smarter. So? If “science” means to you building clever intertemporal optimization models, then it shouldn’t be too surprising that for all one’s intelligence, one has little useful to say about the economy.

15. ssumner
26. May 2009 at 03:44

AdamP, Yes I did read it, and went back to check Cochrane. Here is the exact quote from his conclusion:

“First, our old friend MV=PY with constant velocity (‘stable money demand’) and long and variable lags seems a likely casualty.”

That is exactly my view. Of course with “any velocity specification,” as you put it, the equation becomes a tautology. Obviously it is a matter of opinion as to whether the equation is useful. I doubt that many economists would deny its usefulness for long run cross-sectional studies of M, P, and Y. I am well aware that most economists don’t find it very useful for U.S. business cycle analysis.

Nick and Bill. I have never been interested in the fiscal view, but for what it’s worth, here is my interpretation:

They correctly assume that the long run budget constraint implies the present value of all futures deficits and surpluses (in real terms) must equal zero. I don’t know if I have that exactly right, but there is some such constraint. Then they correctly assume that both monetary and fiscal policy influence the overall budget balance. Then they correctly assume that the future path of fiscal policy implies a defined future path of monetary policy. Then they incorrectly assume that this identity has interesting implications for countries like the US.

In the US, the Fed does pretty much whatever it wants, with little or no consideration of the budget deficits, and this basically forces Congress to accommodate the Fed. In countries with hyperinflation, the central bank is the handmaiden of the fiscal authorities, and then the fiscal view becomes more useful.

26. May 2009 at 03:54

Scott,

I’ll repeat what I said to Bill yesterday.

doesn’t section 3 argue that MV = PY can’t make sense of the current situation with or without constant V?

Here’s a quote from page 5 in section 3:

“In short, financial institutions didn’t want more money and less bonds. They wanted more of both, and less of other assets, and in massive quantities. The “special” or “liquidity” services we usually associate with money applied with nearly equal force to all government debt to these actors.

MV(·) = PY does not really allow us to address this sort of event.”

Yes, Cochrane backs off in the conclusion but in section 3 he’s quite clearly saying that he thinks MV() = PY is not helpful. Of course you can disagree with him but what he says in section 3 is pretty clear and much stronger than what he says in the conclusion.

17. Nick Rowe
26. May 2009 at 04:28

Bill, Scott: Oh, I totally agree; there are lots of problems with the fiscal theory of the price level. I was just writing a post on it, and took a break.

“What is the price level when M = 0?” But I confess I hadn’t thought of that one Bill, LOL! I must use it.

But future fiscal policy still matters. You have only identified one mechanism Bill.

18. 123
26. May 2009 at 12:14

“You keep saying “volumes” are affected, and I still don’t see why that would greatly distort the risk premium. Slightly maybe, but if it moved too far it would bring a lot more speculators in.”
Yes, most of the time speculators would quickly correct the distortion. But I prefer a system that works without a glitch even when hedge fund financing is suddenly curtailed.

“I don’t understand your point about profiting from Fed mistakes. If (today) I expect NGDP growth to be less than 5%, how do I profit? Everyone knows it will be less than 5%-it’s priced into assets.”
Well you actually wrote “On the other hand last September and October it was obvious to me that NGDP growth was going to come in well below target. I might have waited a few days to see if the market jumped in, but if they didn’t I would have grabbed the wheel and yanked it to the right with all my strength, or should I say all my dollars. We’ve all complained about Fed policy, wouldn’t it be nice to be able to put our money where our mouth is. ”
My point was that there were plenty of markets where you could put your money where your mouth was. Did you adjust your portfolio in September and October?

19. Bill Woolsey
27. May 2009 at 02:15

If changes in M _always_ lead to changes in V that leave MV unchanged, then I don’t think the formulation has much use. But if some methods of changing M sometimes impact V, even in an offsetting manner, then the explanatory apparatus is valuable. It is just that those special conditions need to be taken into account.

In other words, increasing the money supply by purchaing zero interest bonds is probably not very effective. V will decrease to offset the increase in M.

But that doesn’t exhaust the possiblities of changing M.

Of course, if you assume V is constant, then this sort of analysis is ruled out.

Cochrane, no doubt, is hearing the same thing from “free market” economists as the rest of us. Base money, M1, and
most importantly, M2, are all rising. Inflation is about to hit as soon as the variably long lag runs out.

Cochrane is correct (and he has said this before, as has Lucas) that open market operations in T-bills are unlikely to
be effective under current conditions.

None of this means that his fiscal model of the price level is any use.

Suppose we had a gold standard with privately issued banknotes and deposits. Price level determination would have nothing to do with future fiscal surpluses. It would only impact the price of government bonds.

Suppose that the government nationalizes the issue of banknotes, keeps a gold standard, and deposits are still private. If the monetary authority is independent of the rest of the government, then the price level has nothing to do with future fiscal surpluses. It would only impact the price of government bonds.

Suppose the goverment has nationalized the issue of banknotes, deposits are still private, and the gold standard is replaced by index futures convertibility on nominal income. The monetary authority is independent. Then the price level has nothing to do with future surpluses. They determine the price of government bonds.

Of course, if the government has fiscal difficulties, it might strip the monetary authority of independence print up money and pay the debt. This is a policy decision. To prevent this, there should be constitutional restrictions on the monetary authority, and so, it should be a constitutional matter. But.. Cochrane’s model treats this policy decision as if it is already made.

When Cochrane says that the “real bills” doctrine holds (according to his model,) then that tells us that the model is wrong. It is one of those long lines of models where “backing” is what is determining the price level.

Where does he go wrong? Money isn’t demanded as an investment, but rather to spend. I accept it because others will accept it. Money doesn’t exist for Robinson Cursoe, and it is going to be impossible to model for a representative agent. You are left with Patinkin’s market vs. individual experiments.

People accept money even if they don’t plan to hold it.
This is a disequilibrium phenomenon. If all you look at
is equilibrim states, then you ignore this key element of
money. (Inside money _is_ inflationary.)

How could anyone claim that the “price level” is determined
when the money supply is zero? The barter exchange rate
between government bonds and various goods? There is no
money price level without money. We can use anything we want as numeraire if there is no money and we are looking at
a barter economy.

Suppose the expected future surplus falls, so that Cochrane’s “price level” rises. The monetary authority
contracts the money supply. (Let us suppose they sell off govenrment bonds.) In reality the price level doesn’t rise.
Suppose that it sells off all of its bonds? Once the money supply is at zero, it can’t contract any more… then…

There is no longer a price level. So, what happens? We return to barter?

If you ignore the essential qualities of the medium of exchange and treat money as an investment asset and look only at equilibrium states, you are going to have trouble.

Nick says that it is a puzzle that the price level is infinite if the budget is expected to be balanced forever.
Well, if the budget is to be balanced forever, then the government debt will never be repaid. And so, it will have zero value.

But this assumes that money only has value if it can be paid off by the issuer with… (money?) In reality, its value will fall if the demand for it is expected to fall and there is no way to reduce the quantity to match the decrease in demand. Cochrane’s approach is to assume that the demand will be expected to fall to zero. Because, people hold money not because they intend to spend it but rather as an investment–like it is stock in the government.

Cochrane’s model also shows that if the government is expected to have deficits forever, the price level is negative. This, is, of course, possible. People with goods pay buyers to take them away. Right.

Money has value because it servces as the medium of exchange. Just because it is easier to model equity claims of a firm doesn’t mean that equity claims on firms are like the medium of exchange.

Government bonds don’t serve as medium of account. They have prices in terms of the medium of account.

People holding govenrment issued banknotes and banks holding deposits at the Fed are not the residual claimants of the U.S. government.

20. ssumner
27. May 2009 at 05:55

Adam, In an earlier paper Cochrane “disproved” the fiscal multiplier by assuming velocity was a constant. Krugman and DeLong mercilessly ridiculed him for this error. But here’s the thing, in the very same paper he later relaxed the velocity is constant assumption and showed that fiscal policy could work. Krugman and DeLong ignored that part of Cochrane’s paper. Cochrance made the mistake of not qualifying his initial velocity is constant assumption, but simply stating is if it were a fact. I defended Cochrane in a post where I mentioned how he had relaxed that assumption later in the same paper. So the bottom line is that by focusing on where Cochrane ends up in the conclusion (which is where most people put in their final word) I am doing him a big favor, I am not pointing out the unrealistic assumptions he may provisionally entertained along the way. I hope you don’t want me to follow Krugman and DeLong, and quote Cochrane saying velocity can be assumed constant, and thus that fiscal stimulus cannot work?

BTW, Congratulations on the new blog. I look forward to saying “did you even read my comment” in your comment section. 🙂

Nick, Yes, That was a good post. You thought of a lot more ways that fiscal policy could work than I would have imagined.

123, The point you miss is that one can only profit from Fed errors if the Fed offers to trade contracts at the target price. Imagine that there had been a private NGDP market in operation last September and October. Could I have profited from my perception that the Fed would miss its target? No, because the market also felt that way. Indeed the ONLY reason I felt they’d miss their target was because the market also felt that way. The NGDP futures contracts would have shown near zero expected NGDP growth. Only if the Fed promised to buy or sell unlimited contacts at a 5% premium, could I have profited from their loss of credibility. Right now the US stock market is essentially where it was around October 10th (S&P) The market had already priced in policy failure. No unexploited profit opportunities in any of my posts—not one.

Bill, You make a number of goods points against the fiscal view. I especially like your discussion of fiscal policy under a gold standard. As long as you are pegged to gold (or any hard currency) expected fiscal deficits won’t move the price level.

Regarding the Cochrane discussion of swapping cash for zero interest bonds, I think we need to stop thinking in terms of disembodied monetary actions, and start thinking of monetary regimes. What is the impact of swapping \$1000 cash for a zero interest \$1000 T-bill? There is no answer to that question without context. What is the monetary regime? What is the target? Is the cash injection expected to be permanent? If so that would raise the price level expected to occur when the economy was no longer in a liquidity trap, and hence depress real interest rates today. Even Krugman would accept that view. But what about if the cash injection was expected to be reversed in 15 minutes? Well then even the most hardline monetarist would say it would have no impact on current asset prices. So my response to the common debates about this question is that we are asking the wrong question, we need to think in terms of alternative monetary regimes.

21. Nick Rowe
27. May 2009 at 07:02

Bill:
“Nick says that it is a puzzle that the price level is infinite if the budget is expected to be balanced forever.
Well, if the budget is to be balanced forever, then the government debt will never be repaid. And so, it will have zero value.”

No. Like the Fiscal theorists, you have forgotten seigniorage. (Except you only temporarily forgot it, since everything else you say makes sense.)

Take a simple example: a government with no bonds, no taxes, but which finances government expenditure (deficits) by printing money. For simplicity, assume zero real growth in GDP.

The fiscal theory of the price level, with no bonds, says:

M/P = PV(primary surpluses = PV(-G)

So the price level should be negative!

But remember in this case inflation is positive, so the real rate of interest on money is negative, so the PV formula becomes (if we assume constant real government expenditure):

M/P = PV(-G) = (-G/r) = (-G/-inflation) > 0

22. Bill Woolsey
28. May 2009 at 02:32

Nick:

Thanks for the lesson, but I am a dense student.

Could you please explain this in words?

Anyway, I thought your claim that Cochranes equation implies an infinite price level with a balanced budget to be correct.

I don’t think that is “correct” about the real wolrd. It just seemed to me that if it all adds up to zero on the right, then M+B/P = 0 and so, 1/P = 0.

It always seemed to me that any model of money and prices that treats money as an equity investment in govenrment is going to require deflation and budget surpluses to work.

I grant that a negative divided by a negative is a positive. Still…

Now, if instead, money is valued as a medium of exchange, then there is no problem with treating the quantity of money as being driven by budget deficits.

But maybe I never understood.

23. Nick Rowe
28. May 2009 at 04:12

Bill: You’re not dense, but a very good monetary economist!

I’m not quite sure what you don’t understand, but I will explain the argument in my last comment differently.

Suppose the US federal government pays off the debt, except money, and then stops collecting any taxes. But the Fed keeps going, and every year prints money, hands it over to the newly-printed money to the government, which spends it. So every year the government has zero taxes, positive government expenditure, and so by definition runs a primary deficit (a negative primary surplus).

You and I know that the money could still have value. But what does the Fiscal Theory say? At first glance since the fiscal theory says that M/P equals the present value of surpluses, and the surpluses are negative, the real money supply must be negative, which means the price level must be negative.

Here is how to “rescue” the fiscal theory from drowning in that absurd conclusion:

To keep the math simple, assume zero real GDP growth, 5% inflation, and 5% money growth rate. Assume velocity is constant over time (it might vary with the nominal rate of interest and the rate of inflation, but I’m supposing they don’t vary over time).

If G is real government spending, the real primary surplus is -G (since taxes are zero). To find the present value of a constant stream, we divide by the rate of interest, so the present value of primary surpluses is -G/r. Now, what is r? If we interpret r as the real rate of interest paid on money, then r in this example is -5%. So the present value of primary surpluses is -G/-5%.

So the fiscal theory of the price level says M/P = -G/-5%. Rearranging we get G=5%(M/P). Which makes perfect sense, because we started out assuming that M grows by 5% per year, and the government spends it.

But we needed the quantity theory of money to figure out how to “rescue” the fiscal theory. So it’s not much of a theory if you need to use its main competitor to make sense of it.

The underlying problem is that we need to separate the fiscal authorities long run budget constraint from the monetary authorities long run budget constraint, and recognise that their liabilities pay different real rates of interest. And that money typically has a real rate of interest below the growth rate of the economy, so that a Ponzi scheme in money is stable.

The fiscal theory tries to lump the two liabilities and two budget constraints together, forgetting the fact that the interest rates are different.

24. David Stinson
29. May 2009 at 14:07

Bill:

“In 1956, Yeager wrote about a scenario in which an excess demand for government bonds pushed interest rates on government bonds so low, that the excess demand for bonds at that point shifted to an excess demand for money. Under those circumstances, open market operations (purchasing government bonds with money) exacerbate the excess demand for government bonds, shunting even more of it to money. The increase in the quantity of money will be matched by an increase in demand.”

I don’t suppose that you would have a reference to that paper would you? I have been hunting all over for it but can’t find it.

25. Bill Woolsey
30. May 2009 at 02:29

“A Cash Balance Interpretation of Depression” Southern Economic Journal, April 1956, 438-47

It is reprinted as the first chapter in _The Fluttering Veil: Essays on Monetary Disequilibrium_ Liberty Fund, 1997. George Selgin, ed.

26. David Stinson
30. May 2009 at 10:05

Excellent Bill. Thanks so much!

27. ssumner
6. June 2009 at 12:36

Nick, It’s been a while so forgive me if I am repeating myself, but I have always thought of seignorage as simply a tax. More specifically, money creation is a tax on real cash balances equal to the rate of growth in the money supply. So in my view the problem is that we define “budget deficits” in the wrong way. Otherwise, I entirely agree with your analysis.