# Those elusive “transmission mechanisms”

Nick Rowe has a great new post on transmission mechanisms.  Here’s the concluding portion, but you should read the entire thing for context:

If I explain why water in the lake has the same level everywhere by supposing it were higher in the North than South and saying it would then flow downhill from North to South you do not falsify my explanation by showing there is no empirical evidence it ever was higher in the North than in the South. I am explaining why there is no empirical evidence it ever was higher in the North than in the South.

The two stories of the transmission mechanism reflect the two equilibrium conditions. The first story assumes that the second equilibrium condition always holds, and explains why the first equilibrium condition holds. The second story assumes that the first equilibrium condition always holds, and explains why the second equilibrium condition holds.

That simple model had two equilibrium conditions and two stories of the transmission mechanism. In a model with n equilibrium conditions there would be n stories of the transmission mechanism. Or more than n, if you combined them. How many equilibrium conditions are there in the real world?

There are not two competing theories of the monetary transmission mechanism in that very simple model. There is no such thing as the monetary transmission mechanism. A “transmission mechanism” is not a causal chain; it is an explanation of why one equilibrium condition holds. The search for the transmission mechanism is a snipe hunt beloved by the people of the concrete steppes. It’s a form of category mistake, like searching for which building on campus is the university.

Let’s apply this to some debates that often arise at this blog.  Consider the following two mechanisms:

1.  Assume that when the base is increased (exogenously and permanently) the public’s preferred MB/GDP ratio (the Cambridge k) doesn’t change, or at least doesn’t change enough to offset the base increase.  In that case NGDP must rise. Also assume wages and prices are sticky.  In that case RGDP also rises.

2.  Assume that wages and prices are sticky, and asset prices are flexible.  In that case a sudden rise in the base may boost some asset prices.  This might lead people to buy more real goods and services.

So which is right?  Both.  This isn’t an either/or situation.  The second mechanism (call it New Keynesian) implies the first (call it the hot potato effect.)  And the HPE implies people have an incentive to buy more real goods and services, when combined with the sticky price assumption.

Update, I notice that Tyler Cowen just linked to a critique of economics that is roughly 1/3 good points and 2/3 nonsense.  Here’s one example of the latter:

17. They think that the world behaves as if their assumptions are true (or close enough).

The first link goes to a Nick Rowe comment:

Consider this hypothetical:

Suppose the central bank has inflation on target (or thinks it does), and output at (what it thinks is) potential output. Then all of a sudden banks start creating loans and deposits out of thin air. If the central bank lets them do it, aggregate demand will increase past potential output, and inflation will rise above target (or, at least, the central bank thinks it will). So the central bank must stop them creating loans and deposits out of thin air. The central bank will raise its rate of interest by whatever it takes to stop banks creating loans and deposits out of thin air. It is exactly as if the banks were reserve-constrained and couldn’t create money out of thin air. Whether you think of the central bank targeting the price (interest rate) on reserves or the quantity of reserves doesn’t make any difference. These are just two different ways of telling the same story about the central bank controlling the supply *curve* (or supply *function*) of reserves. You can think of “supply” as quantity as a function of price and other things, or its inverse, price as a function of quantity and other things.

If the central bank is keeping inflation on target (or thinks it is), the only way that all the banks can expand is if they persuade some people to save more so they can lend more to other people to invest more. Because if banks try to create loans and deposits out of thin air the central bank will stop them (unless of course the central bank wants them to do this because it thinks inflation would fall below target otherwise).

If you have trouble with this, it’s like saying you can think of OPEC setting a price and letting the market determine Q, or setting a Q and letting the market determine P.

Now compare the quality of Nick’s comment (which seems absolutely correct to me) to the second link, which returns to the exact same blog (“Unlearning Economics.”)

There is an obvious problem with Friedman’s snooker player analogy: the only reason a snooker game is interesting (in the loosest sense of the word, to be sure) is that players play imperfectly. Were snooker players to calculate everything perfectly, there would be no game; the person who went first would pot every ball and win. Hence, the imperfections are what makes the game interesting, and we must examine the actual processes the player uses to make decisions if we want a realistic model of their play. Something similar could be said for social sciences. The only time someone’s – or society’s – behaviour is really interesting is when it is degenerative, self destructive, irrational.

It’s hard to disagree with the claim that if, by assumption, nothing explained by economic models assuming rationality is at all interesting, then economics is not very interesting.

Tags:

49 Responses to “Those elusive “transmission mechanisms””

1. Geoff
26. October 2013 at 10:33

Nick’s thinking is too rigidly devoted to equilibrium concepts, which prevents a deeper understanding of the market process.

The economy is never in equilibrium.

2. dtoh
26. October 2013 at 11:13

But Scott, there is a causal transmission mechanism… and it is not the HPE. Consider the following analogy.

A subway system is operating at less than full potential ridership. The local MTA (who operate the subway system) decide they want to issue more subway tokens because they think it will increase ridership. They rule out just giving the tokens away (i.e. a helicopter drop) and instead decide to sell some tokens at a lower price. After doing so they note that ridership has increased. In fact it turns out that there is a very close correlation between ridership and the supply of outstanding tokens. The MTA (thrilled with their success) claim not only correlation but also causality….. a greater supply of tokens causes an increase in ridership. When asked if maybe the increase in ridership was due to the fact that they had lowered the price of the fare (i.e. the price of a token had dropped relative to the price of dollars), the MTA responds by saying “Well maybe that had something to do with it, but the main mechanism is the HPE.”

3. Mike Freimuth
26. October 2013 at 11:18

I’ve heard a lot of complaints about “equilibrium concepts” but I’ve never understood what these people would replace them with. What I’ve never heard is an economist claiming that every market in the real world is always in equilibrium. But if you don’t start by identifying what an equilibrium would be given certain assumptions, then how can you generate any kind of interesting prediction?

4. Mike Freimuth
26. October 2013 at 11:24

dtoh,

People buy subway tokens because they want to ride the subway. People hold base money because they want to use it for transactions. The HPE is based on people using money for the purpose it actually serves. You have created an “analogy” based on a good which serves a totally different purpose which does not lead to a HPE but this doesn’t refute the HPE hypothesis in any way.

26. October 2013 at 12:37

My favorite post by UnlearningEconomics is this one:

http://unlearningeconomics.wordpress.com/2013/08/17/market-monetarism-jumps-the-shark/

In particular the topic of endogenous money and Granger causality came up. I’ve done some googling and apparently the endogenous money enthusiasts have been blogging about the empirical evidence supposedly supporting endogenous money for the past several months and it apparently has gone to their heads. Most of this research involves Granger causality.

I’ve been toying with Granger causality this year, specifically a technique I read about on David Giles blog invented by Toda and Yamamato. I’ve also noticed in ZIRP episodes with monetary base expansion (QE) normally there is corresponding broad money expansion which the endogenous money people say is impossible.

I’ve run Granger causality tests on monetary base, M1, M2. MZM with NGDP over 1959Q1 through 2013Q2. All of the results show bidirectional causality as expected. In fact all of the monetary aggregates are cointegrated with NGDP.

Normally the central bank elastically provides base money at the policy interest rate it chooses to target. If it wants to provide less it is always at liberty to raise the policy rate. The policy interest rate is the instrument by which the central bank determines the level of money supply and the level of nominal GDP. However, at zero interest rates the central bank can no longer change the interest rate to influence the level of money supply. Hence it would be important to determine if QE Granger causes money supply.

I did my analysis with the monetary base and the broadest measure of money supply for Japan, the US and the UK during ZIRP (0.1%, 0.25% and 0.5% for Japan, the US and the UK respectively). For Japan this means their L measure of money supply between March 2001 and May 2006. For the US this means MZM from December 2008 to July 2013. For the UK this means M4 from April 2009 to May 2013. I haven’t tested Japan since the Great Recession since it really didn’t do any significant QE until April 2013.

In Japan’s case monetary base Granger causes money supply and money supply Granger causes monetary base at the 5% significance level each. So I suppose the result is ambiguous although the two variables are obviously significantly correlated. In the US the monetary base Granger causes money supply at the 10% significance level but money supply does not Granger cause the monetary base. Similarly in the UK the monetary base Granger causes money supply at the 5% significance level but money supply does not Granger cause the monetary base.

When I alluded to this at UnlearningEconomics, it was pooh poohed because of all the “empirical evidence supporting endogenous money”. Of course, if a central bank is setting interest rate targets one expects loans to Granger cause the monetary base, and broad money, or the broad money multiplier, which is what some of these studies find (almost all of them are published in the Journal of Post Keynesian Economics). But under QE the monetary policy instrument is no longer the overnight interbank interest rate, something that so far has not dawned on most of the endogenous money enthusiasts.

PeterP first brought up the issue of endogeneity in comment #37 and that thread is strictly a conversation between me and Peter P.

Then at comment #67 Unlearning Economics chimes in:

“In any case, even with a clear correlation I would have this down as an issue of reverse causality, which PeterP has pointed out. I adhere to endogenous money theory, where high income results in loan expansions, and the central bank accommodates the growth in the money base. This means that income and reserves will soar at the same time, but not that the latter is causing the former. Further, it is not inconsistent with the idea that the central bank or government placing restrictions on the issuance of bank reserves will cause problems, as happened in the US in 1937, in Japan and so forth.”

And that thread is a conversation strictly between me and UE (although Philippe threw in some unnecessary snark at a much later date).

Then at comment #86 Unlearning Economics made his/her “last comment” to which I did not respond:

“This causality issue is really the crux of what’s going on. I actually don’t know of any studies that found causality from the money base to the broader money supply; only Kaldor, Kydland & Prescott and numerous endogenous money papers which find it the other way. As I’ve said: the endogenous money story is not inconsistent with there being some effect on the economy from a large MB expansion, as this will make the interbank market more liquid. However, there is a saturation point.”

This last point seems to concede that monetary base expansion may somehow have an effect on the economy and hence on broad money through bank lending. But UnlearningEconomics also seems to be asserting that QE only has direct influence on reserve balances, and no direct influence at all on deposits.

I would like to continue that conversation at a future point because:

1) I’ve done much further econometric analysis on the US QEs that in my opinion conclusively refutes the idea that QE has no effect on broad money and bank lending. What I find is that since December 2008 the monetary base Granger causes loans and leases at commercial banks and that the M1, M2 and MZM money multiplier all Granger cause loans and leases (but neither is the other way around). This is exactly the opposite of what Accomodative Endogeneity (e.g. Basil Moore) predicts.

2) I’ve done considerable reading on the issue since, and the Kaldor and Kydland & Prescott papers that he cites need addressing. Kaldor presents no econometric evidence whatsoever in any of his papers/books. And the 1990 Kydland & Prescott paper used an econometric technique on money that has been used this way five times since, twice by Kydland and Gavin (1995 & 2000), twice by Cooley and Hansen (1995 & 1997) and once by Bergman, Bordo and Jonung (1998). In particular Kydland’s later papers reach far less extreme conclusions, and Bordo and Jonung are old school Monetarists who obviously don’t agree with the central premise of endogenous money theory.

P.S. I’ve had conversations with Nick Edmonds, who I think it’s safe to say considers himself a Post Keynesian. He is neither surprised nor disturbed by my results. So it would be unfair to say all people symathetic to Post Keynesian Economics are like UnlearningEconomics.

6. ssumner
26. October 2013 at 12:55

dtoh, Your example actually proves my point. The cause of the increase in ridership could be seen as either the greater quantity of tokens sold at market prices, or the lower market price. Tow sides of the same coin, er . . token.

Mike, I agree.

Mark. I find it painful to read that endogenous money stuff, I’m glad Nick has taken the time to show the fallacy involved. As for Granger “causality” it doesn’t really establish causality in the normal sense of the word. The “cause” can occur after the effect. Of course what’s really happening is that expectations of the cause produce the effect. But Granger causality misses that.

26. October 2013 at 13:23

Scott,
“As for Granger “causality” it doesn’t really establish causality in the normal sense of the word.”

I would go further than that in saying that there is probably *no* econometric technique that establishes causality in its strict metaphysical sense. But that’s neither here nor there, as we are talking about econometrics, not philosophy.

“The “cause” can occur after the effect. Of course what’s really happening is that expectations of the cause produce the effect. But Granger causality misses that.”

Moreover, if both X and Y are Granger caused by Z, one might still conclude X Granger causes Y. Yet manipulation of X might not change Y. However, the Toda and Yamamato technique offers the opportunity to run the test with three or more variables (provided one actually knows ahead of time which variables to test).

8. Martin
26. October 2013 at 13:24

This discussion reminds me of one of my favorite passages in Samuelson’s foundations:

“It is sterile and misleading to speak of one variable as causing or determining another. Once the conditions of equilibrium are imposed, all variables are simultaneously determined. Indeed, from the standpoint of comparative statics equilibrium is not something which is attained; it is something which, if attained, displays certain properties.”

I guess the notion of an equilibrium is simply a difficult concept to get ones head around.

9. Mike Freimuth
26. October 2013 at 13:45

“I guess the notion of an equilibrium is simply a difficult concept to get ones head around.”

I concur! Much more difficult than most people appreciate. I would like to see critics of equilibrium try to come up with some kind of “disequilibrium model” that isn’t either an equilibrium model in disguise or just a set or arbitrary unconnected conjectures.

10. jknarr
26. October 2013 at 13:47

Scott, as long as we are dwelling on the “underpants gnome” transmission/second step between base growth and NGDP growth, I think some discussion of debt is in order.

Price transmission is questionable when there are two very different components of the monetary base -reserves and currency – that are not universally tradable. Could the effect of preexisting high levels of debt on reserves be the backbone of zlb/liquidity trap observations?

Bank regulations are also instrumental in suppressing currency base money growth in the US. It’s no wonder that currency goes abroad – it has higher utility outside of US bank regulations that restrict currency volumes.

If banks don’t deploy reserves as much due to preexisting leverage, and banks don’t allow currency to circulate in volume, then how does NGDP rise, exactly?

11. Dustin
26. October 2013 at 14:48

I find the snooker analogy to be unhelpful – the CB is not in competition against lending institutions. While snooker is a zero-sum game, banking is not.

“The only time someone’s – or society’s – behaviour is really interesting is when it is degenerative, self destructive, irrational.”

This statement creates a poor reflection of it’s author.

12. dtoh
26. October 2013 at 14:51

Scott,
No. You’re making the same mistake as the subway barons. There can be no increase in the quantity of tokens unless the MTA lowers the price. People hold the amount of tokens they need to ride the subway. The MTA can not increase the issuance of tokens without lowering their price relative to dollars (i.e. assets).

Ridership increases solely because the price of a fare falls in dollar (asset) terms. And… NGDP increases because the price of real goods and services falls relative to the price of financial assets.

You and Nick keep thinking of the money supply in terms of an exogenous increase in the supply, i.e. a helicopter drop or gold discovery. This is not how monetary policy works.

13. jknarr
26. October 2013 at 15:20

Here’s your money, assets, and NGDP model, subway barons.

http://research.stlouisfed.org/fred2/graph/?g=nLV

NGDP appears to slow when assets are relatively high priced to money. When assets are low priced to money, NGDP accelerates.

14. Jim Glass
26. October 2013 at 18:09

> “The only time someone’s – or society’s – behaviour is really interesting is when it is degenerative, self destructive, irrational.”

“This statement creates a poor reflection of it’s author.”

Indeed. Someone is too glib, or if serious has a distorted view of reality. Examples of individual excellence in performance and careers are commonly fascinating and educational. A reason why so many biographies are written and read.

As to societies, the story of the Rise of Rome — all the things that wee little town did *right* while growing in power and influence to construct so much of civilization and contribute to our world still today — is far more interesting than the story of its long, slow, decline, IMHO, and certainly more constructively, positively, useful to us.

15. lxdr1f7
26. October 2013 at 19:16

“1. Assume that when the base is increased (exogenously and permanently) the public’s preferred MB/GDP ratio (the Cambridge k) doesn’t change, or at least doesn’t change enough to offset the base increase. In that case NGDP must rise. Also assume wages and prices are sticky. In that case RGDP also rises.”

You cant assume this. The empirical data certainly doesn’t support this. Creating base doesnt automatically result in a specific amount of gdp just to maintain a ratio. There is many variables affecting how much gdp you get for an increase in base. In part it depends on how that new base affects lending and deposit creation and where those new loans are directed at creation to determine how much gdp you get for an increase in base. If you want to make this assumption then your idea works fine but this is not how the system works.

16. lxdr1f7
26. October 2013 at 19:27

“2. Assume that wages and prices are sticky, and asset prices are flexible. In that case a sudden rise in the base may boost some asset prices. This might lead people to buy more real goods and services.”

It should lead to a weak or insufficient increase in demand for goods and services because a large swath of the population little or no assets and minority of the population hold the majority of assets. The poor that don’t have assets wont increase their spending and the rich wont either. The increase in asset prices will also have a weak effect on the credit channel because people with few or no assets wont become more accessible to credit and the rich with many assets only represent a minority of the population.

17. Saturos
26. October 2013 at 19:28

Greg Ip on the Wicksellian rate of interest:

It’s easier to make a judgment on whether the Fed has it right by looking to the future when the Fed has finally started tightening and the zero lower bound no longer binds. The Fed has already said it expects real rates to be zero in 2016, when unemployment is virtually back to its natural level of around 5.5%.

http://www.economist.com/blogs/freeexchange/2013/10/are-real-rates-too-high-or-too-low

18. Don Geddis
26. October 2013 at 20:54

dtoh: “You and Nick keep thinking of the money supply in terms of an exogenous increase in the supply, i.e. a helicopter drop or gold discovery. This is not how monetary policy works.

An owner of a Treasury bond decides to sell on the open market. He was about to sell to another private individual, but instead winds up selling to the Fed (which is doing an OMO). The Fed takes ownership of the T-bill, and credits the bank account of the original owner (via simple fiat).

Why do you not consider this an exogenous increase in the money supply?

19. lxdr1f7
26. October 2013 at 23:05

Don Geddis

“An owner of a Treasury bond decides to sell on the open market. He was about to sell to another private individual, but instead winds up selling to the Fed (which is doing an OMO). The Fed takes ownership of the T-bill, and credits the bank account of the original owner (via simple fiat).

Why do you not consider this an exogenous increase in the money supply?”

Your example is endogenous. But if the seller sold the treasury to a commercial bank in exchange for money credited in a deposit account it would be endogenous. Some money in the system is endogenously originated and some is exogenously originated.

20. lxdr1f7
26. October 2013 at 23:06

I meant your example is exogenous in my previous comment.

21. J
26. October 2013 at 23:37

dtoh,

You said: “There can be no increase in the quantity of tokens unless the MTA lowers the price.”

Similarly, in the money market, there can be no lowering of the price unless the quantity is increased. The Fed does not set interest rates like the MTA sets prices. The Fed buys and sells in the market until the market price changes. The quantity and price must move together.

22. Unlearningecon
27. October 2013 at 03:14

Hi Scott (and others),

A few things: first the base money stuff. You will not be surprised to know that I find talking to NGDPTers about this as tiring as you find talking to endogenous money proponents, so I’ll be brief.

This is not a cournot market; it’s much more complex. Aiming for p and letting q vary is not the same as aiming for q and letting p vary, which is simply demonstrated by the fact that the latter did not work at all in the 1980s. Nobody has actually disputed the mechanics of banking as described by endogenous money proponents so sometimes I have a hard time seeing the problem: banks makes loans, this expands M1, CB accommodates this demand for money with limited influence. Somebody once helpfully compared this to an electricity supplier.

Also, you say that everything is ultimately determined by expectations. This seems both trivially true and quite wrong. Ultimately we have to ask where expectations come from – which is surely from the reality of what’s happening around people. You might say a credible central bank, but this credibility also has to be established through concrete results, right? So we’re sort of back to the whole transmission mechanism hoo-ha.

You sort of sarcastically reference the fact that I linked back to my own blog a few times in that post. The reason I did this is because, unlike Auld’s list, I wanted to offer an argument for why I think those things were wrong alongside the evidence that economists use it. Some of the links were by me, others weren’t, but obviously some of the points I’d made would have been most explicitly made by me before.

Anyway, aside from navel gazing, let’s talk about the ‘as if’ piece. It is worth noting that saying sports players calculate things can be approximately right in certain situations, but will be wildly off the mark in others:

Simply using their actual decision making processes yields better results, and this is also true for economics.

In any case, the analogy is not that important – I just attacked it because Friedman used it so often. I then went on to talk about actual situations where a slight deviation from assumptions can have enormous impacts. It seems to me that the mechanics of banking are a major area where this is true. I’ve tried to explain why before:

http://unlearningeconomics.wordpress.com/2012/09/27/exogenous-and-endogenous-money-room-for-reconciliation/

Thanks.

23. dtoh
27. October 2013 at 03:46

Don, you said;

“An owner of a Treasury bond decides to sell on the open market. He was about to sell to another private individual, but instead winds up selling to the Fed (which is doing an OMO). The Fed takes ownership of the T-bill, and credits the bank account of the original owner (via simple fiat).

Why do you not consider this an exogenous increase in the money supply?”

When the Fed engages in OMP, they are not replacing another private individual as the purchaser. To the extent that OMP are not offset by an increase in ER (something the Fed can easily control), then OMP will result in a marginal increase in net sales of financial assets by the non-banking private sector. So it is not a substitute purchase, it’s an additional purchase.

J, you said;

“”There can be no increase in the quantity of tokens unless the MTA lowers the price.”

Similarly, in the money market, there can be no lowering of the price unless the quantity is increased. The Fed does not set interest rates like the MTA sets prices. The Fed buys and sells in the market until the market price changes. The quantity and price must move together.”

Not sure if you’re agreeing or disagreeing. Normally the Fed sets the quantity of assets to be purchased and let’s the market determine the price through auction. They could just as easily set the price and let the market determine quantity. To expound on the subway analogy, the MTA could also sell additional tokens (at a lower price) through an auction mechanism. The pricing mechanism is just a detail. Either way, what’s happening with both monetary policy and the subway example, is that the MTA (or Fed) is driving increased demand for real goods and services (or subway rides) by lowering the price of those good and services relative to the price of financial assets.

24. Nick Rowe
27. October 2013 at 04:07

Unlearning: “Nobody has actually disputed the mechanics of banking as described by endogenous money proponents so sometimes I have a hard time seeing the problem: banks makes loans, this expands M1, CB accommodates this demand for money with limited influence.”

Apologies for the fisk:

“banks makes loans, this expands M1”

Basically OK so far. That is *one* of the ways M1 can expand.

“CB accommodates”

No. The CB may or may not accommodate. It depends on what the CB is targeting, and whether it thinks this expansion in loans and M1 will help it hit its target or move it away from the target. If the latter, the CB will *not* accommodate.

“this demand for money”

No! The demand for *loans* is *not* the same as the demand for *money*. If I want to borrow an extra \$100,000 to buy a house, I do *not* want to hold an extra \$100,000 on average in my checking account. The demand for loans increased by \$100,000, but the demand for money is unchanged.

But what is incredibly ironic is that you took a quote from me, on a point where I *disagree* with the majority (orthodox?) view in macroeconomics, to illustrate what is wrong with “orthodox” economics!

25. Nick Rowe
27. October 2013 at 04:32

Unlearning: let me try it this way:

Suppose “Fred” says: “The quantity of money is endogenous and determined by the quantity of money demanded at the rate of interest set by the central bank”.

Fred could be a very orthodox New Keynesian/Neo-Wicksellian macroeconomist.

Fred could be a very orthodox follower of James Tobin.

Fred could be a Post-Keynesian.

Fred could be you (I think?).

There are very few macroeconomists like me who think that Fred is mistaken.

Really weirdly, my views are actually (I think) closer to Steve Keen’s on this question than to the orthodoxy!

Some disagreements cut right across party lines. The simple orthodox/heterodox distinction just doesn’t work.

And it is utterly bizarre that you take my (unorthodox) views on *this* question as an example of what is wrong with orthodox economics!

26. ssumner
27. October 2013 at 05:07

Mike and Martin, Good points.

jknarr, I’ve covered that many times in other posts. Whether NGDP rises depends on the monetary policy of the central bank. Is the base injection permanent, etc.

dtoh, You said;

“There can be no increase in the quantity of tokens unless the MTA lowers the price. People hold the amount of tokens they need to ride the subway.”

The term “need” should never be used in economics. There is no such thing as “need”, only varying quantity demanded at varying prices. If the base exogenously increases then NGDP will rise until there is “need” for the larger base. That’s the HPE. If the quantity of subway tokens increases then their price falls until there is that much “need” to ride subways.

lxdr, You said;

“You can’t assume this. The empirical data certainly doesn’t support this.”

I can’t imagine what data you are referring to. I’ve studied money my entire life and the data certainly does overwhelmingly support this claim.

Saturos, Good Ip post.

UnlearningEcon, You are wrong about the P and Q question, you are confusing a policy issue with a theoretical issue. No one claimed anything about Q being held constant, the claim was that setting a P and setting a Q are equivalent at a point in time. Of course if you hold either one constant, the two policies will diverge over time. Indeed in Nick Rowe’s post he assumed Q was not held constant, but was varied as needed to stabilize inflation. I can’t speak for Nick, but I see Q as the causal mechanism but do not favor targeting Q, so your criticism is off base.

As far as the “correct model” Yes, we’d all like to be mind-readers, but until and unless we can do so we come up with the best models we can. Any economics model that did as well as Friedman’s snooker model would be far ahead of anything we currently have.

As far as banking, the endogenous money crowd (AFAIK) confuse the micro problem of how banking operates with the macro problem of how nominal aggregates get determined. In other words they treat as a single problem changes in the nominal size of the banking system’s balance sheet and changes in the real size of the banking system’s balance sheet. But they are two different problems!

27. lxdr1f7
27. October 2013 at 05:40

“”You can’t assume this. The empirical data certainly doesn’t support this.”

I can’t imagine what data you are referring to. I’ve studied money my entire life and the data certainly does overwhelmingly support this claim.”

MB and gdp havent increased at the same rate

25 fold increase in gdp from 1960 to 2007
16 fold increase in mb over same period

http://research.stlouisfed.org/fred2/graph/?chart_type=line&s%5B1%5D%5Bid%5D=BOGMBASE&s%5B1%5D%5Brange%5D=10yrs

http://research.stlouisfed.org/fred2/graph/?chart_type=line&s%5B1%5D%5Bid%5D=GDP&s%5B1%5D%5Brange%5D=10yrs

What data do you refer to?

28. Saturos
27. October 2013 at 05:50

29. Saturos
27. October 2013 at 05:51

Notably, it has been argued that regular movements in the markets reflect a wisdom that transcends the best understanding of even the top professionals, and that it is hopeless for an ordinary mortal, even with a lifetime of work and preparation, to question pricing. Market prices are esteemed as if they were oracles.

This view grew to dominate much professional thinking in economics, and its implications are dangerous. It is a substantial reason for the economic crisis we have been stuck in for the past five years, for it led authorities in the United States and elsewhere to be complacent about asset mispricing, about growing leverage in financial markets and about the instability of the global system. In fact, markets are not perfect, and really need regulation, much more than Professor Fama’s theories would allow.

30. Ralph Musgrave
27. October 2013 at 06:14

Martin’s above quote from Samuelson strikes me as nonsense: in particular Samuelson’s claim that “It is sterile and misleading to speak of one variable as causing or determining another..”.

Take an object that floats in water: it will always return to an equilibrium. That happens because the difference between mass of water displaced and mass of the object, causes the object to move up or down.

But now economists like Nick Rowe are trying to tell us the latter transmission mechanisms are irrelevant. I suggest it is precisely the transmission mechanisms that bring about the equilibrium (or fail to).

31. jknarr
27. October 2013 at 08:07

Scott, we can try to move beyond “not enough” monetary policy tautologies to understand why policy is not working to lift NGDP in the here and now after five years. I think that reserves, debt levels, and currency market repression have something to do with it.

Expectations need to be rooted in experience, I.e. credibility. Markets can form the expectation that the base remains locked up, unable to affect equilibrium prices in the real economy.

If the Fed’s bought rusty nails for a \$100 bill each, from the public, 10 billion times, I am quite certain that it will do more to prices than another \$1 trillion pomo into reserves: this a real utilitarian and theoretical problem, so please discuss.

Consider also that “permanent” might be a different beast in an IOR world.

32. ssumner
27. October 2013 at 09:06

lxdr, You said;

“MB and gdp havent increased at the same rate

25 fold increase in gdp from 1960 to 2007
16 fold increase in mb over same period”

How in the world do you consider that inconsistent with my claim? What did you think I was claiming, that V is constant???

Saturos, You picked the same two paragraphs as I picked.

Ralph, The question of causality depends on one’s perspective.

jknarr, You said;

“Scott, we can try to move beyond “not enough” monetary policy tautologies to understand why policy is not working to lift NGDP in the here and now after five years.”

All you need to do is consider the fact that the Fed viewed it as a “close call” the decision on whether to taper last month. That’s enough to explain it, nothing more needs to be said.

33. Mike
27. October 2013 at 09:34

ssumner

You assumed MB to GDP to be at some static level. And in reality that is inconsistent.
How could the ratio between MB and GDP hold if V isnt constant? (assuming sticky P)

34. J
27. October 2013 at 09:50

dtoh,

My point is that an exogenous change in price that results in a change in quantity is equivalent to an exogenous change in quantity that results in a change in price. Moreover, buying and selling bonds is no different from a helicopter drop (except that a helicopter drop has a fiscal stimulus component). Either way, interest rates and/or inflation must adjust for the public to be willing to hold the money.

35. jknarr
27. October 2013 at 09:55

So the Fed says that they will stabilize base/ngdp. So what? They hit similar base/ngdp levels in 1930-1940 and NGDP went to 10-20%.

The point is that similar expansion of the MOA is having very different macro effects. I don’t think that expectations are sufficiently strong an explanation to explain the difference in similar policies=different results.

Lets not hide behind tautologies in addressing such a fascinating and consequential problem.

http://research.stlouisfed.org/fred2/graph/?g=nMr

36. Dustin
27. October 2013 at 12:31

Jknarr,

Isn’t the point that 1) when MB/GDP increase, 2) NGDP increase also?

Using your graph, this assumption sure seems to hold even if the degree of correlation is variable. While the MB/GDP ratio does change, it does not change enough to offset the base increase (ie, NGDP increases) precisely what Scott detailed as seen below:

“1. Assume that when the base is increased (exogenously and permanently) the public’s preferred MB/GDP ratio (the Cambridge k) doesn’t change, or at least doesn’t change enough to offset the base increase. In that case NGDP must rise.”

37. Dustin
27. October 2013 at 12:36

Nick

“No! The demand for *loans* is *not* the same as the demand for *money*. If I want to borrow an extra \$100,000 to buy a house, I do *not* want to hold an extra \$100,000 on average in my checking account. The demand for loans increased by \$100,000, but the demand for money is unchanged.”

This sounds an awful lot like saying demand for money – the extra avg in my checking account – is analogous to demand for future consumption vice current. This presumes of course that demand for money coincides with the intent to use said many for (one of) it’s purpose(s) as a medium of exchange and eventually exchange it for stuff. Is this a correct reading?

38. dtoh
27. October 2013 at 13:57

Scott,

The term “need” should never be used in economics. There is no such thing as “need”, only varying quantity demanded at varying prices.

Yes of course. That’s what we mean when we say “need.”

If the base exogenously increases then NGDP will rise until there is “need” for the larger base. That’s the HPE.

Then you should rechristen HPE with a new name and stop calling it a mechanism. It’s just a trivial equality. Both “hot potato” and “effect” imply some form of causality. If that’s not how you’re using the word, you should call it something else like the Money Nominal Product Equilibrium (MNPE).

And BTW, the inability to offer a sound and convincing explanation of causality for the efficacy of NGDPLT is IMHO the main reason it’s not more widely understood and accepted by main stream economists and policy makers.

39. dtoh
27. October 2013 at 14:44

J, you said;

“My point is that an exogenous change in price that results in a change in quantity is equivalent to an exogenous change in quantity that results in a change in price. Moreover, buying and selling bonds is no different from a helicopter drop (except that a helicopter drop has a fiscal stimulus component). Either way, interest rates and/or inflation must adjust for the public to be willing to hold the money.”

No. The key thing is not the money. The key thing is the price of financial assets relative to the price of real goods and services. If this changes, then there is an increase (or decrease) in real AD. A helicopter drop does not change the relative price of financial assets and real goods. Thus it will simply cause a rise in the overall price level.

40. J
27. October 2013 at 16:23

dtoh,

41. dtoh
27. October 2013 at 16:35

J
Real aggregate demand or NGDP.

42. dtoh
27. October 2013 at 16:46

J,
Correction real AD or real GDP.

43. Jim Glass
27. October 2013 at 16:50

“No! The demand for *loans* is *not* the same as the demand for *money*. If I want to borrow an extra \$100,000 to buy a house, I do *not* want to hold an extra \$100,000 on average in my checking account. The demand for loans increased by \$100,000, but the demand for money is unchanged.”

This sounds an awful lot like saying demand for money – the extra avg in my checking account – is analogous to demand for future consumption vice current. This presumes of course that demand for money coincides with the intent to use…

It can be put rather more simply: It’s very often mistakenly stated that “the interest rate is the price of money” when in fact it is the price of leasing or renting money. The price of money itself is something different, the amount of goods/services that must traded for it to obtain ownership of it.

Leasing and owning are two different things (whether the subject is real estate, automobiles or money). The demand for one thing is different than the demand for a different thing.

44. J
27. October 2013 at 18:59

dtoh,

Then how do you distinguish between real AD and real AS? In an AD/AS model, we can think of AD as NGDP and AS as the breakdown of NGDP into inflation and RGDP. Monetary policy is used to regulate NGDP, i.e. to provide stability to AD. You correctly point out that a helicopter drop would cause inflation. That’s the point. We can raise NGDP by raising inflation. If a NGDP shortage, through sticky prices and/or wages, is causing a shortfall in RGDP, then raising NGDP through inflation will also raise RGDP.

45. Dtoh
28. October 2013 at 01:01

J
Good question. I tend to simplify and think of inflation as being caused by an expected imbalance in short term supply and demand (plus some hysteresis plus some other stuff). When you have an AD shortfall, OMP will get you more real growth and less inflation. At higher rates of real growth you getter a bigger mix of inflation.

If the Fed targets NGDP, the mix will be self-adjusting (eg if the mix is all inflation then you will have no real growth which will cause an abatement in inflation.)

Regardless though, OMP by causing an increase in the real price of financial assets results in a marginal increase in the exchange of those assets for real goods and services, ie an increase in AD.

46. ssumner
28. October 2013 at 05:30

Mike, No, I never assumed V was “at some sort of static level”. Can you find the quote where I say that?

jknarr, the 1940s are an exception to the rule that low rates correlate with low NGDP.

dtoh. Just because a transmission effect is hard to visualize, doesn’t make it unimportant. It’s also hard to visualize water flowing “downhill” from one side of a lake to another (using Nick’s example.) One doesn’t actually see that happening.

47. dtoh
28. October 2013 at 06:16

Scott,
The transmission mechanism is not hard to visualize, it’s just that the mechanism not HPE.

If you want to know what the mechanism is for a lake having a constant water level, most people expect an explanation having to do with gravity and fluid dynamics. The HPE is like telling them that the mechanism is the “lake leveling effect,” and then…. when asked to explain it, you tell them it’s the mechanism causing water level at one end of the lake to be at equilibrium with the water level at the other end of the lake.

48. dtoh
28. October 2013 at 06:18

Scott,
The asset price mechanism is not only simple to visualize, it’s crystal clear.

49. lxdr1f7
28. October 2013 at 06:27

“MB and gdp havent increased at the same rate

25 fold increase in gdp from 1960 to 2007
16 fold increase in mb over same period”

How in the world do you consider that inconsistent with my claim?”

In the first point you try to assume something that isnt realistic such as a stable mb/gdp ratio:

“1. Assume that when the base is increased (exogenously and permanently) the public’s preferred MB/GDP ratio (the Cambridge k) doesn’t change, or at least doesn’t change enough to offset the base increase. In that case NGDP must rise. Also assume wages and prices are sticky. In that case RGDP also rises.”