August 2014

Sun Mon Tue Wed Thu Fri Sat
          1 2
3 4 5 6 7 8 9
10 11 12 13 14 15 16
17 18 19 20 21 22 23
24 25 26 27 28 29 30
31            
Blog powered by Typepad

« Contributions to Economic Sociology | Main | Peter Klein's PhD Course in Austrian Economics »

TrackBack

TrackBack URL for this entry:
http://www.typepad.com/services/trackback/6a00d83451eb0069e20133f344bcb9970b

Listed below are links to weblogs that reference Krugman's Way-Too-Simple Model:

Comments

Feed You can follow this conversation by subscribing to the comment feed for this post.

Krugman seems to be implying that there is no modeling outside Keynesian orthodoxy, or at least that he's never heard of any such thing. Krugman really is a poster child for demagoguery.

Rajan's column on Bloomberg may not contain sketches (great MS Paint skills by the way, Paul), but his papers usually aren't shy with models...

"It's the consistent application of a theory that's been around longer than yours."

This is no argument either. It's just a claim to vain authority grounded in a years count? Might be that Austrian economics has some valid explanation for the current mess. I don't think so. Unfortunately it has no valid answer to come out of the malaise. Or to paraphrase Harry Hopkins: "People don't eat in the long run, they eat every day."

"The fact that with heterogeneous capital goods, even given idle resources, one cannot assume that C and I will move directly together and not opposite each other never enters the picture."

Speaking of heterogeneity, all the talk of "cash" on corporate balance sheets is a perfect example of how C and I correlate in strange ways in these strange times. You've likely heard the bullish arguments about US corporations sitting on tons of "cash." In fact, as several commentators have pointed out, this "cash" is really debt in the form of corporate bonds. Moreover, this cash isn't really on the sidelines at all. It's being parked in other debt instruments like Treasuries, which in turn fund our "stimulus" spending.

So, corporations can borrow in the bond market at 3-7% (give or take) and park those funds in Treasuries which earn far less. Corporation are generally NOT using those funds for investment and new hiring. (Healthy balance sheets???) Moreover, it's not as if Treasury itself is "investing" in sustainable production processes, but unsustainable "stimulus" projects and transfer payments. Still, as you point out, "it's all part of AD or Y," so what's the worry?

@ Steven Horwitz

I think you are a bit to harsh on Krugman. Take your monetary equilibrium view (which I share): Think of a fall of velocity. This is, ceteris paribus, associated with high short-run real rates of interest AND with lower aggregate demand (effective money supply < aggregate output). You don't need capital to show the AD-slope above.

But Krugman's point can be attacked from another angle, completely from within the believe system of modern macro. For liquidity traps are associated with zero lower bound iff there is no other way for the monetary authority to increase effective money supply (Mv). Helicopter Ben suggested that there is always a way to do so. I do not understand, why Krugman persistently ignores this point of view.

Best

Raghu Rajan's model can be found in this paper http://faculty.chicagobooth.edu/douglas.diamond/research/Diamond%20Rajan%20Interest.pdf that he co-authored with Doug Diamond which explicitly references ABCT. It is however atleast an extension of the Austrian theory if not a significant restatement as I argue here http://www.macroresilience.com/2010/08/03/raghuram-rajan-on-monetary-policy-and-macroeconomic-resilience/.

To borrow from my post: "The conventional Austrian story identifies excessive credit inflation and interest rates below the “natural” rate of interest as the driver of the boom/bust cycle but Rajan and Diamond’s thesis identifies the anticipation by economic agents of low rates and “liquidity” facilities every time there is an economic downturn as the driver of systemic fragility. The adaptation of banks and other market players to this regime makes the eventual bust all the more likely."

In my opinion, this is a pretty accurate statement of the Greenspan/Bernanke era and the impact of the "Greenspan Put".

Forget AE for a moment.

I have been re-reading papers by Brunner and Meltzer, which cover the monetarist revolution. In essence, Krugman wants to refight all the issues settled in those debates. Krugman's model isn't as sophisticated as IS-LM. B/M sliced and diced the latter for leaving out securities.

Let's take one seminal contribution of Milton Friedman: the permanent income hypothesis. That explains why stimulus doesn't work. The Bush tax rebates retested and reconfirmed the hypothesis. Nonetheless, the supporters of stimulus say "push on."

Krugman's starting point, about a presumed inverse causal relationship between unemployment and real interest rates, is belied by empirical evidence, namely that for much of the last three decades (to choose one period), higher real rates were accompanied by lower unemployment rates. A better theory, such as Steve adumbrates above (and explains in detail in his other published work), would explain why.

Stephan writes:

Might be that Austrian economics has some valid explanation for the current mess. I don't think so. Unfortunately it has no valid answer to come out of the malaise. Or to paraphrase Harry Hopkins: "People don't eat in the long run, they eat every day."
-----
Okay, what's wrong with the Austrian theory?
And as far as pointing out that an alternative, the Austrian approach, has been around longer, Steve was simply stating a fact of which Krugman seems blissfully ignorant.

@Stephan,

Two points

(1) You have misapplied the economic concept of the long run. It is not a period of time. Mario Rizzo has explained this very nicely in posts at ThinkMarkets.

(2) "It [AE] has no valid answer to come out of the malaise." What is your cure for a hangover? (Hint: there is none.)

The central thrust of AE (which, in this respect, is classical economics) is that a crisis and a collapse in asset prices (the present malaise) is the inevitable result of a mania (asset boom).

The adjustment involves lots of repricing (changes in relative prices). What is your valid answer -- other than allowing prices to adjust? No change in nominal magnitudes can solve a problem of incorrect relative prices.


Jerry O'Driscoll;

"In essence, Krugman wants to refight all the issues settled in those debates. Krugman's model isn't as sophisticated as IS-LM."

That is consistent with his "How Did Economists Get It So Wrong?" plan for macroeconomics; scrap the last 80 years or so and start anew from the General Theory...

Steve - We need more of this discussion when you appear on FOX.

Stephan,

> This is no argument either. It's just a claim to vain
> authority grounded in a years count?

The length of time that a theory has hung around for doesn't give it more likelihood of being correct.

However, it does show that the theory has been created in an ad hoc manner to answer another theory, and that's the point here. Steve's point is that Austrian Economists didn't invent ABCT in order to dispute Krugman or Keynes' views. ABCT is older than their theories.

Jerry O'Driscoll,

> Krugman's model isn't as sophisticated as IS-LM

That was exactly what I thought.

> Let's take one seminal contribution of Milton
> Friedman: the permanent income hypothesis. That
> explains why stimulus doesn't work.

I'm not really sure I agree about this though. In Monetarism the question comes down to if the agent has good-quality expectations (be they adaptive or rational) or not. I had a discussion with Barkley about this, and after reviewing it I think he's right. We can't really assume that the common household does have good-quality expectations.

I think that households and small businesses face high costs in forming expectations. Only large businesses, banks and speculators can really do that well. Small businesses must rely on their accounts, and households must rely on heuristics. This point is closely related to Pietro M's post earlier about the beginnings of ABCT.

None of this really means that stimulus must necessarily work though.

Great comments, Jerry. Thanks for that.

BTW, Krugman says, "That is, other things equal, demand is higher, the lower the real interest rate. Do you really want to quarrel with that?" I guess I might an issue, actually. Considering his axes, I guess that's supposed to be an IS curve, but then he speaks vaguely of "demand." Anyway, he needs an LM curve or and AS curve depending on whether his downward sloping curve is supposed to be an IS curve or some sort of AD curve. Instead he's got one half of the model and a full-employment output. Is he mixing up what Patinkin called individual and overall experiments? What precisely does he have in mind? I don't find his model so very clear, really.

Roger is right about the confusion in his model.
Re: higher demand/lower real interest rates, about the only thing lately for which there has been a sustained increase in demand has been bonds. Certainly not labor, capital goods, and bank loans. Serveral publicly-traded purveyors of consumer goods took hits recently. All this while real rates have drifted downward. Presumably, an increase in rates will come about the time the economy turns up, or maybe a bit later if the economic upturn anticipates the rate increase(s).

I agree with Krugman.

He is claiming that the natural interest rate is negative. Could be. If so, then negative real market interest rates are necessary for saving and investment to be equal and markets to clear. (The only way that the market interest rate could be too high is if there is an imbalance between the quantity of money and the demand to hold money. His proposal to increase expected inflation would make money more costly to hold, reduce the demand to hold money, and bring it into equilibrium with the quantity of money.)

Horwitz is assuming that the natural interest rate is.. well, I don't know. But anyway, that the lower real market rates proposed by Krugman would be below the natural interest rate where ever it happens to be.

If the real market interest rate is above the natural interest rate, then a lower real market interest rate can expand both consumption and investment. I have no idea why anyone would think that the interest rate elasticities of the elements of those parts of aggregate real expenditures are somehow wrong. The demands for all sorts of things could expand, clearing up shortages.

And yes, if we imagine that there are already bottlenecks creating shortages of everything that people want to buy more of, then Krugman is wrong. There is no shortage of money. The market interest rate is not above the natural interest rate. Could be.

I must admit that I favor a stable growth path of money expenditures consistent with a stable price level more or less. (3%) I don't favor targeting a higher inflation rate so that near zero nominal interest rates on the shortest and safest financial assets are consistent with more negative real interest rates on those assets. I think growth path targeting would change expectations and raise the natural interest rate on even short and safe assets. If that doesn't work, then I suppose the second best alternative is for the Fed to purchase longer term and riskier assets. But the first best solution is for nominal interest rates on short term and safe assets to turn negative if necessary to clear markets. Exactly why should people earn positive real yields on T-bills due to mature in one month? If you want that taxpayer guarantee, then maybe you should pay.

These worries about malinvestments always fail to take sufficient account of the yield curve. Why can't interest rates on short term and save assets be very low, and interest rates on longer term and riskier assets be higher, providing a proper signal of the long run balance between the forces of thrift and productivity?

Bill, in your post above in what sense are you using the term "natural interest" rate. There are several definitions which aren't quite the same.

the lower real interest rates are supposed to clear up surpluses.

"These worries about malinvestments always fail to take sufficient account of the yield curve. Why can't interest rates on short term and save assets be very low, and interest rates on longer term and riskier assets be higher, providing a proper signal of the long run balance between the forces of thrift and productivity?"

No one disputes that a normal yield curve has an upward positive slope. That's not the issue.
The issue is, what would it be like under free banking vs. the system we have now?
While short rates on safer fixed income assets might well be very low (whatever that is), I doubt they'd be as low as they are now. After all, the gub'ment and its "independent" central bank have well-documented institutional incentives to drive them down and keep them there, until Comrades Bernanke, Geithner et al. shout out the all clear signal, no doubt long after the ship has weighed anchor and left the port.

Professor O'Discoll: The permanent income hypothesis gives us a lot, but I don't think it necessarily disproves the positive effects of a stimulus. If we are below our average projected income level over our lifetimes, we're not going to save more today in order to pay off the taxes tomorrow. Or perhaps I'm mistakenly conflating your views on permanent income with the Ricardian Equivalence?

Professor Woosley: I don't understand the concept of a negative interest rate. Do you see anyone pleading to give you a dollar in exchange for $0.97 a year from now? If that's the case, I would love to get in that business. If you want to say that interest rates are something besides that, I don't know of any microfoundations that would support it.

Even if we were to accept Krugman's framework, I don't understand why the real interest rate doesn't go down asymptotically to zero instead of cutting across the zero bound.

Ryan:

Negative real interest rates are quite common.

The real rate of return on zero interest currency is the negative of the expected inflation rate.

A variety of financial assets can have negative real returns as well, nominal returns less than the expected inflation rate.

Now, negative nominal interest rates are possible if a "dollar" that you might lend takes the form of a deposit account that bears a negative nominal interest rate as well. If you assume that all such deposits are redeemable in zero nominal interest bearing currency, then the lower bound of for the interest rate is the cost of storing currency.

Do you have some deeper microfoundations in mind?

The natural interest rate is the level of the interest rate where saving and investment are equal. And that is the same as the interest rate where real expenditure (consumption and investment in the simple case) equals the productive capacity of the economy.

> The natural interest rate is the level of the
> interest rate where saving and investment are equal.
> And that is the same as the interest rate where real
> expenditure (consumption and investment in the simple
> case) equals the productive capacity of the economy.

That's basically Wicksell's view isn't it?

In "savings" would you include money held?

You think that the first condition: saving & investments being equal, necessarily implies the second. But, for the rest of us that's not necessarily the case. This is one reason why I find your posts difficult to understand sometimes.

In what part of the post above are you referring to the natural interest rate in the first sense and in what parts are you using the second sense?

I understand that structural abnormalities can give rise to negative real interest rates. Perhaps example of trading a dollar today for $0.97 tomorrow was misleading.

I was speaking only to the natural interest rates you mention. I agree that the natural rate of interest is the rate that clears the market. But I don't see this necessarily means that we need to identify the level of inflation which would set supply equal to demand; I see supply as representing the disutility of not using a dollar today and the demand as the marginal product of new investment (etc). Why would these factors only intersect at a point below zero? Perhaps these are just deeper microfoundations, but without some story like price stickiness, I don't find Keynesian analysis is useful insofar as it can be useful. Can you suggest any deep parameters that are consistent with my suggested interpretation of the natural rate of interest, or offer a reason my interpretation is inadequate?

Stepp:

Suppose the interest rate necessary to clear markets next year are lower than the interest rate necessary to clear markets over the next 5 years.

Suppose purchasing a capital good that would be profitable with the interest rate necessary to clear markets over the next year, if they were to persist, would be unprofitable if the interest rate was at the level consistent with clearing markets over the next 5 years.

Isn't the answer having a low one year interest rates now and higher 5 year interest rates?

In other words, it isn't an "upward sloping yield curve" that is the point, but rather shifts in the shape of the yield curve based on what interest rates need to do to provide the correct signals and incentives to balance consumption and investment over time.

Consume this year and implement investment projects that pay off this year. Don't start investment projects that pay off in 5 years or expand consumption for 5 years.

It isn't the "normal" shape of the yield curve, but the changes in its shape.

Surely, the market response to excessively low short term interest rates in a steeper yield curve?

As for your theory that central banks are all about keeping interest rates low at all times--I don't agree with that. New Keynesian theory and central bank practice is focused on making sure interest rates are increased enough to keep inflation from rising.

We all know that there have been monetary cranks who believed that an expansionary monetary policy could permanently lower real interest rates and shift the distribution of income away from capital and towards labor. Those of us here know that many of them were libertarians. In my opinion, Mises was a bit too focused on confronting that fallacy. But imagining that central banks are always seeking to permanently seeking to lower interest rates is a mistake.

are always seeking to permanently lower interest rates..

OK.. back to the JI budget.

@Bill:

Thank you for presenting your side of the argument in a respectful and cogent manner.

I have two questions:

1) It seems like you're ascribing a great deal of importance to the fact that markets aren't currently clearing. Why? The fact that markets aren't clearing suggests there are exogenous factors (such as uncertainty) preventing the move towards a neoclassical equilibrium. Perhaps economic agents are being irrational; perhaps they're behaving perfectly rationally. Why are our expectations any more valid than theirs?

2) Why do you support a stable price level? Prices are signals; they convey information about scarcity and wants. Using monetary policy to mask relative price changes could keep economic agents from discovering some very important information. Isn't that type of monetary policy simply trading short-term stability for a long-term increase in the economy's signal to noise ratio?

Sorry for the double post; that should read "decrease in the economy's signal to noise ratio."

Ryan,

Permanent tax reductions have permanent effects. Temporary tax rebates (or spending) have small, transitory effects. An individual will rationally spend the interest rate (discount rate) x the principal amount. Say $10 out of $100 (roughly in line with past experience).

Permanent changes lead to immediate changes in behavior, and individuals will discount the future to the present. The economic long run beigins kicking in immediately.

I'm not trying to defend a stimulus, but I don't believe I agree with the reasons you gave. If the government exchanges dollars sitting in bank vaults for treasury bills and spends the money on purchasing roads and digging ditches, does the construction worker employed spend $10 of his salary and put the rest in the bank? Are financial markets efficient/prescient enough to coordinate the rest of the economy so something analogous effectively happens. I have problems believing that's the case... there's too much Lucas in it. I don't doubt that the line of reasoning that says we can stimulate the economy by the mechanism I described is fallacious, but for different reasons.

Ryan,

This is basic price theory, and deals with individual behavior. It has nothing to do with Lucas or macro. If you announce a program is temporary, then ordinary working folk do not act as if it is permanent.

Standard macro treats people as idiots. They may not be prescient, but they are not idiots. To confuse a one-time payment of $100 with a permanent annual fund of $100 does require people to be idiots.

Woolsey writes:

"It isn't the "normal" shape of the yield curve, but the changes in its shape.

"Surely, the market response to excessively low short term interest rates in a steeper yield curve?"
---
Agreed with the first statement, but not necessarily the second. In fact, Fed policy can affect the term structure of the curve over its entire length. When short rates fall, wouldn't "rational" investors search for higher yields by investing further out on the curve? That in turn would push rates there lower.
It's called interest rate arbitrage, and it's alive and well, despite what Comrade Greenspan thinks. As Exhibits A-D, look no further than the (sub-market) discount rates (Buffett uses a 10-year bond rate) used to value cash flows of intestments that throw them off-stocks, bonds, private equity, and real estate in the run up to the most recent asset crash and burn.

I don't think I said that central banks keep interest rates low (i.e., sub-market) *at all times*, or seek to keep them low *permanently*-- at least I didn't mean to say that.
But central banks do have a proclivity to lower them, and these actions have been implicated in asset booms. The test is would they be as low under free banking, which has a much better historical record of producing a stable economic environment?

I don't think it is correct to call Krugman's diagram a "model." It is just a geometrical statement of his assertion that we need negative real interest rates to get to full-employment.

Steve,

I find Krugman dislikable as an economist and person. I have no particular desire to defend him personally.

But his models are "way-too-simple" for what purpose? I get the impression that most of the complexity you accuse Krugman of ommitting concerns factors that he has explicitly rejected as causes of the ongoing recession.

If one believes that the recession is entirely a product of a monetary shortage, falling nominal expenditure, inadequate aggregate demand, or whatever else one may call it, then a model that accounts for the complexities you describe is likely *too* precise for Krugman's problem situation

Moreover, Krugman seems to be more concerned with illustrating a principle than giving a qualified and precise description of the situation. He even calls his model the "quick-and-dirty version."

Precision for precision's sake does not a solution to a problem make.

Professor O'Driscoll,

Basic price theory tells us that the Robbinsian maximizer will behave in the way you described. Naive Keynesianismn and Monetarism treated people like idiots, yes. New Keynesianism doesn't; it just makes ridiculous assumptions. And I thought I remember reading in your book that you and Rizzo felt affinities for Post-Keynesianism, though I don't have the book in front of me so please excuse me if I'm putting my foot in my mouth there.

Even granting a Robbinsian maximization of a temporary tax cut (which, given the behavior of people who win the lottery and countless others, I do not think is realistic), that leaves aside several things. It does not explain why workers who get a job from a stimulus will somehow treat that income differently than other income and save ninety percent of it. It does not explain the legitimate times when the Ricardian equivalence breaks down even for the Robbinsian maximizer (e.g. imperfect bequests). I know that if given $100,000, I would spend more than $4,000 of it this year. Is it really such an uninformed position to suggest that real people, when confronted by the dark forces of time and ignorance, will not behave "optimally" as dictated by price theory, upon receiving a check for a few thousand dollars?

The negative real equilibrium (natural) interest rate is a new keynesian deus ex machina to describe what hey can't understand.

I did some "research" on the Japanese crisis months ago and several papers came out. One was by Eggertson and Woodford, and it was about monetary policy at zero interest rates. There were others which were more specific about Japan, I believe, but I don't have a bibliography at hand.

Premise: in new-keynesian models of the "neowicksellian" type, there is a real interest rates that mimics the flexible-price behavior of the economy even under sticky-prices, i.e., it assures closeness to full employment. Monetary policy steers the economy toward full employment without having any adverse effect except possibly inflation, because being a f(K,L) model there is no chance of introducing some notion of structural distortions and unsustainability in this worldview.

How do new-Keynesian justify the Japanese lost decade? They don't, they claim without proof that real equilibrium interest rates have been negative for a decade, or two. This hypothesis is required for them to fit the data, but it has no justification, exactly like the real business cycle hypothesis that some unobserved gigantic negative productivity shock caused the Great Depression.

New-Keynesian applying their theories to Japan are stuck between two unappealing possibilities: the first is to assume the unlikely (a ten-year long negative equilibrium real interest rate), or get rid of their hyper-simplified models and talk about structural unsustainability. They preferred the former alternative, preferring professional survival to scientific relevance.

Hello Pietro,

you write: "The negative real equilibrium (natural) interest rate is a new keynesian deus ex machina to describe what hey can't understand. [...] New-Keynesian applying their theories to Japan are stuck between two unappealing possibilities: the first is to assume the unlikely (a ten-year long negative equilibrium real interest rate) [...]."

I share your critical view on New Keynesian models (DSGE with rational nominal rigidities). But the negative real rate is a more general concept, applying to all theories which incorporate a Fisher relation. If you have a central bank reducing the nominal rate to zero, and you have some inflation, the real or natural rate is negative.

The case of Japan, however, is approached differently. Here we had almost zero nominal rates and deflation. Thus, the real rate is positive. For Krugman and others, this is the problem: real rates were to high, i.e., monetary policy overly restrictive.

For Krugman this is a liquidity trap such that monetary policy is ineffective. He asks for fiscal stimulus.

For Bernanke, Woodford, etc, this is sign of a policy failure (because they see that transmission channels related to interest rates are not exclusive). BoJ was simply to restrictive or conservative.

Central banks do not control AD primarily by interest rates, but by announcement effects that trigger inflation expectations. Bernanke suggested price level targeting, which assures an more than proportional increase in inflation expectations as soon as the price level falls below its target.

Scott Sumner on The Money Illusion makes this point (over and over again ;)).

Wow! This discussion was already well under way.

I just wanted to point out a problem with a claim of yours, that speaks to what you said the other day about Krugman "arguing in good faith". You wrote:

"At the NYT today (thanks Bill Stepp), Krugman has a post defending what he calls his macro model and arguing that folks who oppose easy money/negative real rates/raising expected inflation are guilty of ad hoc arguments to justify their "predjudice" (which is presumably coming from somewhere irrational) against easy money."

He did not say that people who oppose that stuff are guilty of ad hoc arguments. He said that Rajan was and he said that many others are. You read him too harshly which is probably why too many of you think he's never arguing in good faith. Reread the post - he's pretty clear that a subset of people argue without good theory.

Arash Molavi:

if the central bank sets interest rates to zero and there is inflation, then the *market* real rate of interest is negative, not the natural one.

Creating inflation and reducing nominal target rates is a way to achieve a negative cost for funds on the interbank market, so that it can be set equal to the mysterious *natural* real rate of interest which assures full employment.

Now, *market* real rates can be persistently negative if there is price inflation but little Fisher effect on nominal rates, as in Eastern Europe, and now more or less all around the world.

I questioned the reasons of assuming negative *market* real rates as a credible assumption for an economic model. My impression is that they need to *assume* it otherwise their theories couldn't explain the facts regarding Japan.

This is equivalent to saying "then a miracle occurs"... what is, in fact, a persistent negative *natural* real rate of interest? Is an economy in which equilibrium requires *lenders* to pay *borrowers* to accept their funds, which is a blatant absurdity in terms of temporal preferences. Now, this may happen if there are nominal rigidities, but it's insane to assume this can happen for more than a few months. It is even more insane to believe that negative real *market* rates can be an adequate policy to insure a meaningful operation of the market process. But New-Keynesians don't care about it.

comment, part 2.

Has been, or is, the BoJ and the Japanese government restrictive in their monetary and fiscal policies?

This is nonsense.

What has happened in Japan from a monetary point of view is that nothing could have prevented a reduction in nominal spending because the monetary multiplier system (the banking system) totally broke down.

What has happened in Japan from a fiscal point of view is that no amount of fiscal debt has been sufficient to jump start the economy. Some people might say - and they have (Kuttner and Posen on the Brookings papers) - that that policy was not enough active. However, this hides the real question: if 200% debt/GDP ratios are contractionary, what is an expansionary ratio? 1,000? 10,000? It simply doesn't make any sense*.

I'll set aside fiscal policy because I normally consider it irrelevant or harmful, and there is no surprise in the failure of these policies in Japan as far as my economic worldview is concerned.

What about monetary policy? The money supply is the product of high-powered money and the monetary multiplier. The former is controlled by the central bank, the latter is largely endogenously determined by the profitability of investments, their riskiness, the rate of risk aversion, and the state of the balance sheet of the banks. If some of the endogenous causal factors go astray, central banks can't do nothing by simply moving the monetary base or fixing the interbank rate, or monetizing long-term assets, or whatever.

Under these conditions, the monetary channel is broke, monetary policy is ineffective, and the whole story of keeping aggregate demand from falling becomes mere wishful thinking. Solutions? There are four: a recession (and consequently a banking/financial crisis), bypassing banks, rescuing banks, or deflationary stagnation (Japan). I prefer the former one, at least it doesn't destroy the market system in the long run.

I add that rescuing banks helps in inflating the money supply only if the return rates on investments are high enough to enable banks to loan up their new funds. If the crisis was a purely bank crisis, like equity losses on leveraged positions, it can work. If the problem is a general overpricing of capital goods relative to consumption goods, it can't work.

So, you are left with chronic stagnation (Japan's preferred policy) and bypassing banks, i.e., creating an enormous amount of inefficiency and corruption by socializing credit markets. This is not really an option.

* They could have argued that 200% debt/GDP ratio wouldn't have been reached if fiscal policy had been expansionary since the onset, but this is begging the question of fiscal policy effectiveness. The problem is that if the effectiveness of fiscal policy requires evergrowing deficits to keep the economy in motion, in the long run we will be dead, killed by keynesian policies. The real question is how a 5%, 10% or 15% deficit can be insufficient, and what's to be done when fiscal policy is constrained by solvency issues.

Pietro,

thank you for your detailed comments. I respond shortly to one aspect of your statements above. You write right at the beginning:

"if the central bank sets interest rates to zero and there is inflation, then the *market* real rate of interest is negative, not the natural one."

True, but note that in modern literature the natural rate is a long-run concept applying to a situation where all own rates of interest are uniform across markets. Own rates of interest are exchange ratios, perfectly analogues of Arrow-Debreu prices.

In other words: own rates are market rates. Since the natural rates is defined for uniform market or exchange rates, its is also a market rate. What you have in mind when speaking of a natural rate, is Wicksell's original concept, the marginal productivity of increasing roundaboutness. But

1) this is not the concept used in the modern literature. When Krugman, Bernanke, Woodford say natural rate, they mean intertemporal exchange ratios. So in applying the Böhm-Bawerkian concepts may lead to a misunderstanding of what these guys actually want to say.

2) whereas the productivity aspect of course remains important, it is a determinant of the natural rate, not the rate itself. In all standard models you find productivity growth as a major exogenous variable co-determining the natural rate as an intertemporal price relation.

Best

Ryan Murphy & Jerry O'Driscoll,

I agree with the microeconomic view that Jerry gives. But, as Ryan says I don't think it really deals with the problem. Take the UK for example, before the recession the tax rates vs spending rates were unsustainable even if there had been no recession. Gordon Brown's government ran deficits all the way through the boom. In that case how can the normal household know that a portion of their income is temporary. I don't think they can.

Arash & Pietro,

Great discussion. In the thread above I asked Bill Woolsey to describe his version of what a "natural rate" is. He gives two definitions which are those that Wicksell gave.

There are multiple problems here with definitions. In Wicksell (as far as I understand it) he has a theory linking the two meanings of natural rate. If the savings-investment market clears then full employment results.

If we doubt this link then we have to doubt the other link, that a rate of interest that fulfils one of the two criteria above also results in a sustainable production structure.

I'm not sure I understand Arash's last definition of natural rate, but it sounds like another incompatible one.

Arash Modavi:

I'm not convinced.

As far as I know, Neowicksellian models do define a natural rate, and they define it as the rate of interest which assures full employment, so that the market rate shall be forced to be equal to it thanks to the central bank's policy.

The market is in a recession when the market rate is higher than the natural rate. Short run divergences are the engine of cyclical fluctuations, whereas in the long run there are no cycles to talk about, i.e., it would be growth theory.

Without nominal rigidities, the natural rate would be spontaneously achieved, and a real business cycle model would ensue. With nominal rigidities, however, market rates shall be steered by the central bank in order to mimic the efficient behavior that the market cannot achieve.

It is true that there is only one market rate, this is a defect of all mainstream macroeconomics, in which there is no distinction between return rates on investments and money rates on loans. But this is another question: in neo-wicksellian models, the natural rate is not a market rate, in fact it is also an unobservable variable which shall be estimated.

Cfr:

"The natural rate is defined to be the equilibrium real interest rate that would
obtain in a ficticious replica of the economy in which nominal adjustment is complete. [...] To
be clear, the natural rate is not defined to be a long-term real interest rate; it is a short-term rate, defined period-by-period, and with a long-run central tendency that can also (slowly) shift over time."

BIS working papers:
http://www.bis.org/publ/work171.htm

I haven't read this paper, but normally BIS papers are better and more reliable than my blog comments, I don't know why. :-)

Molavi, not Modavi, I'm sorry about the mispell.

Pietro,

workhorse DSGE models have a single equilibrium rate because the have a single commodity! There are no own rateS, just one of them. Thus, it is possible to tell some short-run dynamics "as if" there is a single equilibrium rate. This doesn't hold true for n-commodity worlds (n>1). Further, if you have an incomplete-market economy such that Arrow securities take over intertemporal and risk adjustments, you introduce uniform short-run risk-free rates. That's true. But in consequence, future spot prices have to take over the coordination task (in contrast to present futures prices as in Arrow-Debreu economies). Then you have to introduce Ratex (REE) to ensure general equilibrium. BUT EVEN IN THIS CASE, THE SHORT-RUN EQUILIBRIUM RATE OF INTEREST IS AN EXCHANGE RELATION, and thus more than the marginal productivity of increasing roundaboutness. Think of Wicksell's NEUTRAL rate.

Current,

thanks for your comment. I think we have three Wicksellian definitions of the natural rate:

1) the rate balancing saving and investment,
2) the rate equal to the natural rate (marginal productivity of increasing roundaboutness),
3) the rate that keeps price level stable at an undetermined initial level.

As J. Koopmans, D. Davidson, F.A. Hayek, G. Myrdal, and Selgin, etc. point out, these three hold true for stationary economies. In progressive economies #1 and # 2 together are in conflict with #3.

I don't know if this is of any help.

Best

oh, I meant to say: three definitions if the neutral rate of interest ... sorry.

Arash,

I'd agree with you about that.

> 2) the rate equal to the natural rate (marginal
> productivity of increasing roundaboutness),

But, what does this bit mean here. When you say "natural rate" here what do you mean? A rate that creates a sustainable production structure?

I think you and Pietro are confusing each other. Your certainly confusing me. I think we're stuck in that "Hayekian Triangle" that Steve Horwitz wrote about last week.

Current,

you may be right about us and the triangle ;).

In Wicksell's Interest and Prices the natural rate is the rate that just suffices to sustain the stationary level of (net) income by inducing capital-owners not to consume their (free) capital and instead to leave the subsistence fund untouched.

... oh, and stationarity of course implies structural conservation.

This is the concept which is NOT the one used in New Keynesian models. At least this was my point against Pietro.

I think this has gone beyond my knowledge of the subject.

However, I'll just add one more thing. I think that Pietro is coming to the problem from an Austrian direction, whereas Arash is coming from a more neoclassical direction. When Pietro says that a variable is "unobservable" I think what he means is that it's unobservable in real life, it is a theoretical or modeling concept. When Arash says that natural rate is a price he means that the natural rate defined by definition #1 in his list is a price in a model.

Current:

Yes, something like Wicksell's notion.

Yes, saving includes changes in money holdings, repayments of debts, as well as accumulation of assets.

Most fundamentally, the relationship between the natural interest rate, real expenditure, and the productive capacity of economy follows from the reality that income not consumed is saved.

However, there is an issue of the relationship between saving (and investment) and income. The natural interest rate is the level of the interest rate that keeps saving and investment equal if income is at a level consistent with full employment. (As opposed to some other level of income that might result in a different supply of saving and/or demand for investment.)

I don't think the natural interest rate depends on the "true" marginal value product of additional roundaboutness. Or that only through the perceptions of entrepreneurs.

While I think it is possible to imagine negative natural interest rate scenarios that are not pathological, I think the pathological case is the more likely scenario.

Something bad is expected to happen in the future that motivates more saving and less investment, and so the natural interest rate is lower. If it extreme enough, it is negative.

To say that it takes an negative real interest rate to equate saving and investment is the same thing as saying it takes a negative real interest rate for consumption and investment to add up to the match a level of real income and output consistent with full employment.

None of this is inconsistent with changes in the composition of the demand for various goods impacting the productive capacity of the economy.

And it certainly doesn't mean that the natural interest rate today will turn out to have been correct in retrospect. If the additional saving and reduced investment were based on some delusion, then, in hindsight, the savings and investment decisions would have been in error.

Suppose people foolishly believe that cold fusion is about to replace gasoline powered cars. And so, the demand for gasoline drops off. The equilibrium price is lower, even if this is all a big mistake.

By the way, defining the natural interest rate in such a way that monetary policy based upon interest rate targeting is a mistake. It is a way of looking at monetary disequilibrium (and equilibrium) that can be illuminating. I think it is especially useful with pure credit monies. And with mixed or ambiguous systems (like ours) it has mixed value.

If the demand to hold money equals the quantity of money (evaluated at a level of income consistent with the productive capacity of the economy,) then the market interest rate is equal to the natural interest rate because it clears credit markets. The distinction occurs when the real market rates that clear credit markets result in an excess supply or demand for money and an imbalance between real expenditures and the productive capacity of the economy.

The changes in the purchasing power of money that clear this up in the long run will simultaneously impact the real supplies or demands for credit, and generate market rates equal to the natural rate.

If the natural interest rate is negative, and nominal rates cannot be negative, then either the market process creates expected inflation or else, it impacts the supply of saving (or demand for investment) so that the real market rate equals the natural rate. For example, the current price level falls below its expected future level. The expectation that it will rise again results in a lower real market rate--negative if needed. Or, the Pigou effect results in higher real money balances, higher real wealth, lower saving, and so a higher natural interest rate.

I am not sure how this all fits into Arrow Debreu economies. The real world is complicated enough. And I don't think it is well characterized by A-D types of reasoning. I do think Wicksellian reasoning tells us something.

Saltman:

If there are long lines at the gas station because either buyers or sellers face uncertainty, I think that suggests some kind of market failure. The price should be higher, clearing the market for gas. The role of prices is to coordinate, not break down because of uncertainty.

There are people who want to work and buy consumer goods now. Complaining that the Obama administration is at fault because of uncertainty (say, due to health care policy) is absurd. Uncertainty should effect the prices, but not leave people frustrated by shortages or surpluses.

No one knows how profitable investment will be? Well, people trying to shift consumption to the future at low risk should receive a low return, maybe negative. If they don't want to shift consumption to the future on those terms, then consume. If they don't want to consume, then don't earn income. I'm quitting work because I don't want any consumer goods now and don't want to save because the return is too low, because there is too much uncertainty, is reasonable. Do you think that is what is happening?

The role of prices is to clear markets. If they aren't clearing, prices aren't doing their jobs.

As for price level stability, I do not favor using monetary policy to mask needed relative price changes.

I don't believe that changes in the price level provide any useful signal when the problem is an imbalance between the quantity of money and the demand to hold money. This is especially true with the increase in money demand is temporary. And I sure as heck don't think people who want to earn income to consume now should just wait until the demand for money falls.

On the contrary, the problem is that the changes in prices (or really, the growth of real demand and past trajectories of prices) actually result in changes in particular outputs that would be appropriate if they did reflect changes in the composition of demand but are inappropriate because they "signal" a change in the balance between the quantity of money and the demand to hold money.

With a gold standard, changes in the price level do tell us something about the cost of jewelry and the profitability of gold mining. One benefit in a sea of confusion. With the sort of system we have today, what they tell us depends entirely on rules. The fixed quantity of money or fixed growth rate of money gives us only worthless information (what price level is needed to adjust the real quantity of money to the demand to hold it.)

I favor a target for a growth path of money expenditures. It is the least bad option, in my view. And so, that doesn't mean a stable price level always. The price level actually would vary based upon shifts in the growth path of potential income. And money incomes, in aggregate, would not be stable, but would grow at a rate that reflects the average growth rate of potential income.

I don't pretend that is perfect, but I certainly see this as the least bad macroeconomic environment for microeconomic adjustment. I think the notion that all nominal incomes should be unanchored and then shift whatever amount needed to keep real expenditures equal to productive capacity is a really, really bad macroeconomic environment for microeconomic adjustment. But the regime I favor would not result in a stable price level at all times, just on average over time.

I think if some nominal interest rates sometimes need to be negative to keep the expected value of money expenditures to keep growing on the targeted growth path, the figure out out to get those nominal interest rates negative. Anyone who takes that to mean that they can build a dam because real interest rates will be negative for the next 50 years is a fool. And they will lose money on any investment based on that foolish notion. Think about it.

If people begin panic buying of gasoline because they think it is all going to disappear next year based on peak oil, and the price of gasoline shoots up to clear markets, and someone builds a refinery with a 20 year lifetime that is only profitable at those high prices, they are going to lose money when people realize that gasoline is not going to run out next year.

One other thing, that isn't directed at anyone particular--"balance sheet recessions" are nothing but the paradox of thrift. If people have too much debt and pay it down they are saving. The natural interest rate is low (ceteris paribus) and the real market interest rate needs to fall enough to clear markets, keep saving and investment in balance.

Are there any countries which has experimented with negative interest rates? I have heard that Hong Kong tried that approach once, and if it's true it would be good if someone could elaborate on what came of it.

Sweden set slightly negative nominal rates on reserve balances that banks hold at the central bank. I think the Fed should follow a similar policy at this time. (I think they should float the interest rates paid on reserves, and keep it a fix number of basis points below the yields on short term government bonds. So, that would be negative now, but presumably positive once the economy recovers.)

Bill,

I read through this whole thread again to make sure I understand it. Looking back, I think your remarks are basically correct. It could be that the natural interest rate is now negative.

However, I think the real problem is expectations of future inflation. Investors are unwilling to invest in low-return investments now because they believe that rates could rise soon. The problem is that if the central bank supply more than the necessary quantity of money then price inflation will result and through the Fisher effect longer-run rates will rise too. I know this is similar to Keynes' view, but I'm coming around to the view that he was right about this to some extent. The issue that brings this about the low natural rate is an exogenous one that comes from the central bank monetary system.

I don't agree with you that a flexible interest rate could adjust to any such uncertainty. As you point out yourself the underlying problem is the demand for money not the price of debt. So, we should think of the problem in those terms. If the government cause uncertainty that changes the demand for money then can a free banking monetary system easily adjust to that? I don't think that it necessarily can, because it's a quantity adjustment not a price adjustment. Free banking, or your expenditure rule, could never remove the negative effects of uncertainty shocks entirely.

"Nor does this enable us to see how the negative real interest rates of the boom of the last decade gave us tons of malinvested capital from which we are still trying to extract ourselves. "

I'm sorry to bring this up again but I'm just not getting what I think is a clear answer. What malinvestment? We've already seen that despite the boom in home prices, the actual number of homes is not all that out of line considering population growth and depreciation of existing homes.

We also worked through hypotheticals concerning investors buying pizza ovens only to discover the future economy doesn't want pizza or ranchers investing in buffalo in anticipation of a future mozzarella cheese boom that doesn't materialize. While all these decisions are bad for the people who make them what is clear is that the price for such 'foolish investment' is born in the period it was made. The person investing in a pizza oven in year 1 only to discover no one wants to buy pizza in year 2 'pays' the cost not in year 2 but year 1. If he had a time machine he could go back to year 1 and have enjoyed more consumption.

But in year 2 the price of the oven has already been paid and it's not 'malinvestment' but positive investment. In year 2 the economy can enjoy more pizza. OK maybe the economy doesn't value pizza as much in year 2 because of changing tastes but pizza just becomes real cheap and the market clears. Maybe the original oven owner lost the oven in bankruptcy but his loss is a gain for a new investor who picks up a new oven for next to nothing.

The only example I can think of as true malinvestment would be one that in year 2 costs more than it yields in benefits BUT scrapping it would cost even more. The example I came up with was a 'bridge to nowhere'. Say it costs $10B to build in year 1. Going forward its net benefit is only $1M (the fees paid by the few bridge users to ferry services before the bridge was built). But the bridge requires $3M in yearly maintenance costs (and this is mandatory, not spending it will cause the bridge to fall apart and create a safety hazard). Scrapping the bridge in a safe way costs $2.5B.

This would be a valid example of malinvestment IMO. The bridge reduces consumption by $2M a year ($1M benefit netted against $3M cost) but this drag on production would cost much more to scrap so the economy must 'live with it' until its willing to bite the bullet and spend billions to scrap it.

But hardly any investment in the real economy looks like this. The absolute worse case usually is that an investment is worth nothing and it is sold for what little scrap value its metal has or it's dumped in the garbage.

Boonton:

The opportunity cost of the pizza oven is the hammer that could have been produced. While more pizzas are produced than otherwise, those extra pizzas are worth less than the houses that could have been produced.

Of course, the traditional Austrian malinvestment story is the reverse. We made hammers to build houses, and because the hammers are build, we do build more houses, but those extra houses are worth less than the pizzas we could have had if instead of houses pizza ovens had been produced.

The low interest rates made houses, and so hammers, look for valuable than pizza ovens, and that was an error.

At least that is how I understand it.

I think the error in Boonton's logic is here:

> But in year 2 the price of the oven has already been
> paid and it's not 'malinvestment' but positive
> investment

There is an issue here between accounting methods and economic reality. It may be that accounting methods require that the return or loss on a piece of capital equipment are indexed to a particular year. But, the economic effect only becomes evident in the year in which problems become evident to the business owner. Because it's at that time that the owner changes his behaviour.

> Maybe the original oven owner lost the oven in
> bankruptcy but his loss is a gain for a new investor
> who picks up a new oven for next to nothing.

No it isn't. Whomever owns it the value of the pizza oven is the net present value of the future pizza profit it can produce. If it is found that this net present value is lower than was expected then that is a loss that is not compensated for anywhere else in the economy.

Bill,

What you are describing is opportunity cost and I do agree with you but opportunity cost is about comparing alternative universes. In universe A there was investment in pizza ovens on the mistaken belief that pizza would become more popular. In universe B there was investment in other things. Universe B enjoys a higher level of income to universe A. In that sense malinvestment is true just as it's true "I should have been a doctor".

But this isn't true as an explanation of the business cycle which isn't about compareing universes but time periods in the same universe. We have are recession if, in year 2, output is less than year 1. But year 2 shouldn't have less output than year 1. Even those those pizza ovens weren't desired, year 2, all else being equal, has the same output as year 1 plus more pizzas.

To get to a recession, to have real 'malinvestment' you have to move from simply bad investment choices or 'too much' investment. The bridge to nowhere example might work. What also might work is if you invested in ovens at the expense of maintaining some important piece of capital. If in year 2 we had more pizza ovens but lost Google because we didn't bother to keep the servers from falling apart then yea maybe that's 'malinvestment' that might plausibly cause a recession.

But here the cause of 'too low' interest rates seems to fail. If interest rates were really low, why would Google invest in silly pizza ovens rather than their servers? No interest rate zero or above makes the 'birdge to nowhere' a sensible investment.


Current
"No it isn't. Whomever owns it the value of the pizza oven is the net present value of the future pizza profit it can produce. If it is found that this net present value is lower than was expected then that is a loss that is not compensated for anywhere else in the economy."

But the new owner of the pizza oven picked it up at a bankruptcy auction. Clearly his bid would be based on the revised estimate of the ovens present value. That pizza turns out to be less popular than estimated a year ago would already be factored into his choice when he places his bid. He wins the oven and starts churning out pizzas. In terms of whether or not we have a business cycle, year 2 sees the capital put to use making goods.

> But year 2 shouldn't have less output than year 1.
> Even those those pizza ovens weren't desired, year 2,
> all else being equal, has the same output as year 1
> plus more pizzas.

Your forgetting about capital maintenance and replacement. If all capital goods were everlasting then ABCT could cause no recession, only a fall in growth below it's maximum potential. The problem in real life is that the Pizza making equipment has been built instead of making replacement capital that would have been consistent with the overall production structure.

> But the new owner of the pizza oven picked it up at
> a bankruptcy auction. Clearly his bid would be based
> on the revised estimate of the ovens present value.
> That pizza turns out to be less popular than estimated
> a year ago would already be factored into his choice
> when he places his bid.

If such a bankruptcy auction does produce a lower price (and I think it would) then that's only because of the illiquidity of specific capital goods.

The revision downward of the value of the oven occurs whomever owns it.

My point here is only that the value of the pizza oven can't be expect to be as high as it was before.

> He wins the oven and starts
> churning out pizzas. In terms of whether or not we
> have a business cycle, year 2 sees the capital put to
> use making goods.

Yes. But, that only means that we have more goods in year 2 overall if there is little need for capital maintenance. In practice though that isn't the case, most investment spending doesn't go into developing new output, it goes into replacing old output.

During the boom demands are generated for certain goods and services that are not sustainable. Investment funds are pushed into these sectors at the expense of other sectors. Those other activities will not necessarily be maintained. This was the point of my example of the Buffalo mozzerella farmer earlier. He switched to Buffalo from Cows, as such he didn't keep up his maintenance of previous capital.

"Your forgetting about capital maintenance and replacement. If all capital goods were everlasting then ABCT could cause no recession, only a fall in growth below it's maximum potential. The problem in real life is that the Pizza making equipment has been built instead of making replacement capital that would have been consistent with the overall production structure."


Why would a decline in interest rates cause capital owners to neglect their existing, productive capital in favor of investing in new capital with unproven returns? In other words, why would Google let their servers rot in order to buy new condos in fly by night developments just because interest rates are 1 or 2 points 'too low'?

"During the boom demands are generated for certain goods and services that are not sustainable. Investment funds are pushed into these sectors at the expense of other sectors. Those other activities will not necessarily be maintained. This was the point of my example of the Buffalo mozzerella farmer earlier. He switched to Buffalo from Cows, as such he didn't keep up his maintenance of previous capital."

This would imply the recession is a supply problem. Because a lot of useless pizza ovens were built we now not only have pizza making ability we don't want but have less ability to make other stuff because we neglected it while we were building up our army of ovens.

OK but a supply shock should result in stagflation. The stuff we do want should be getting more expensive since there's less supply of it (again because we let the capital rot). Likewise our utilization rates should be high since the capital that wasn't allow to rot now has to try to meet all the demand. Our price levels are falling, though, and our utilization rates are low. Where is the evidence of a shortage of capital?

Perhaps what should be entertained here is a fusion of the theories? Perhaps we see a minor Austrian capital allocation problem masked by a Keynesian demand drop in 'animal spirits'. Because demand is too low we can't see which sectors are really short of capital.

It smacks of totalitarianism in the school system; the student is a commodity and the teacher is the factory line-worker producing shiny widgets. Lucy and Ethel on the chocolate conveyor belt were more successful in their production than a teacher could be based on very flawed set of criteria.

Cafe Carlu is a clean, modern cafe in the Cité de Architecture on the the Trocadero with a small menu of simple,well prepared sandwiches, salads,quiches and plats. So what's the big deal?

The comments to this entry are closed.