« Quantitative Easing circumvents banks' capital constraints to increase M1 | Main | Scott Sumner's Plan for monetary policy »

Comments

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

Hi Nick

I must read up on the textbook answer to this, but I had some thoughts on the same issue in response to a different Paul Krugman column:

http://www.knowingandmaking.com/2009/02/in-real-world-aggregate-demand-versus.html

I saw something recently which alerted me to the idea of drawing the AD curve against inflation rate rather than absolute price level, which makes some sense. And the second-order effect that you mention, the effect of prices changing while stock of wealth stays the same, is the most convincing mechanism to me.

See where it says "URLs automatically linked" by the comment box? Not really...

http://www.knowingandmaking.com/2009/02/in-real-world-aggregate-demand-versus.html

Sorry, looks like it does link after a refresh, or after a few minutes. Feel free to delete the duplicate.

Your posting about antiques made me think quite hard about this same subject. I didn't have any answers but the conversation between you and Scott Sumner over the last few weeks has been very stimulating. Understanding the mechanisms by which money works is a fascinating challenge and presents many of the kind of logical paradoxes that first fascinated me as a mathematician. My hope is that some of these paradoxes can be taken advantage of in the successful management of monetary policy, but it feels like there's such a limited understanding of these phenomena among economists that that won't happen any time soon. It gives me hope that there are a few thoughtful people talking seriously about it. I somehow imagine a series of letters between Locke, Berkeley, and Hume along similar lines, trying out the arguments before they find their way into the official texts.

I've written a review of Niall Ferguson's Ascent of Money over the last few days and I was hoping for a bit more of this kind of introspection in that book. It's not really there, unfortunately, but I guess there's a level of geekery in this discussion which is not entirely audience-friendly.

Hi Leigh!

If you draw the AD curve in {inflation,real income} space, hold the nominal interest rate constant, and assume expected=actual inflation, it slopes upwards (i.e. the wrong way). That makes the long run equilibrium (vertical LR Phillips Curve, upward-sloping SRPC) unstable. In fact, that's a great way to illustrate the problems of using the interest rate as the control instrument.

If you hold the money supply constant, and draw the AD in 3-D space, it slopes down with the price level, but up with the rate of inflation.

I half-finished Niall Ferguson's book. It's good, but I got distracted (I distract easily).

How to build a mathematical model with money as the medium of exchange is not easy! Beyond me. (But then I only got a D in A-level maths!)

The bleeding continues in Alberta. Latest numbers on major construction spending aren't pretty:

April: $234001.2 x 10^6
February: $261735.9 x 10^6
December: $270727.3 x 10^6
November: $279300.0 x 10^6
October: $286527.2 x 10^6

http://www.albertacanada.com/statpub/1120.html

(Note the summary table erroneously still says February - the PDF is April)


Leigh,

the blog post that you link to pretty much completely misunderstands econ101. Their is a reason economists are so hung up on REAL interest rates.

Here's the big quote from the other blog:

"It may be another example of the one-versus-all fallacy. The demand curve for a single product will of course slope downwards (with rare exceptions such as Giffen goods). But this is because, for a given amount of income, we can choose between buying one good and another. Thus, the price of a product in the standard demand graph is the price relative to other products.


In the case of the aggregate demand curve, total demand is meant to be based on an average price level, but a price relative to what? If this describes total demand in the whole economy, what is the substitute good that we are supposed to buy instead?"

The answers to the questions in the second paragraph of the quotation is easy: what do we buy instead of current consumption? Future consumption.

Aggregate demand is based on an average price level, but a price relative to what? Relative to future consumption.

The price of future consumption in one period's time in terms of consumption today is (1/1+r) where r is the REAL interest rate. Raise the real rate and you lower the price of future consumption, hence just as in the one good vs another case, people substitute more future consumption for less present (they save more).

On the other hand, lower the real rate and the reverse happens.

This is why all models, RBC or NK or whatever els, all macro models, feature the real rate as one of the most important variables.

Nick,

I think you're misunderstanding Krugman, he's quite clearly talking about debt deflation and not getting his result from an AS-AD diagram. Here's a quote from the link:

"In particular, falling wages, and hence falling incomes, worsen the problem of excessive debt..."

He does not get his argument from the AS-AD diagram.

Adam:

Am I misinterpreting Paul Krugman? I dunno. Take this passage:

"Here’s how the paradox works. Suppose that workers at the XYZ Corporation accept a pay cut. That lets XYZ management cut prices, making its products more competitive. Sales rise, and more workers can keep their jobs. So you might think that wage cuts raise employment — which they do at the level of the individual employer.

But if everyone takes a pay cut, nobody gains a competitive advantage. So there’s no benefit to the economy from lower wages. Meanwhile, the fall in wages can worsen the economy’s problems on other fronts."

Sounds to me like he's (correctly) making the (valid) distinction between the demand curve for an individual firm's output and the demand curve for all firms' output in aggregate, which is the AD curve in {p.y} space (or maybe {W,y} space. Bryan Caplan makes the same interpretation.

As you say, when we draw the individual firm's demand curve (or the demand curve for an individual good like apples) as downward-sloping, it's because we are holding other prices constant, so the real or relative price of apples is changing. So what is the real thing that changes when we draw the aggregate demand curve, and change the general price level?

The standard "textbook" answer is the real money supply. The nominal money supply is held constant when you draw the AD curve. That's why you get the downward-sloping AD curve from the ISLM, for example.

(Under fixed exchange rates you hold the nominal exchange rate and foreign price level constant, and allow the nominal money supply to adjust endogenously, so it's the real exchange rate which changes. The reason for the downward-sloping AD in this case is exactly analogous to the downward-sloping demand curve for apples.)

Your answer is to hold the (expected) future price level constant (and presumably hold the nominal interest rate constant too), so the real interest rate is changing by definition. Interesting, but it's not the "orthodox" answer.

Bryan Caplan is also just wrong. Here is what he said in the post you link to:

Caplan: "1. Cutting wages increases the quantity of labor demanded. If labor demand is elastic, total labor income rises as a result of wage cuts."

Krugman made quite clear he was talking about wages and prices falling in tandem. I'm pretty sure Krugman has in mind an unchanged real wage. If the real wage is unchanged then so is the quantity of labor demanded. Econ101 stuff here.

Caplan: "2. Even if labor demand is inelastic, moreover, wage cuts reduce labor income by raising employers' income. So unless employers are unusually likely to put cash under their matresses, wage cuts still boost aggregate demand."

The same critique applies here as well, no change in the real wage means no change in anyone's real income.

But suppose we do think that it's just a wealth redistribution, if employers and labor have the same marginal propensity to consume then it's a wash, it doesn't raise aggregate demand. If Caplan is claiming that employers marginal propensity to consume is higher than labor's he should say so and try to back it up because that seems like a tough case to make.

Nick, no you're misinterpreting what I'm saying. And we've already discussed this:#

Nick: "The standard "textbook" answer is the real money supply. The nominal money supply is held constant when you draw the AD curve. That's why you get the downward-sloping AD curve from the ISLM, for example.

(Under fixed exchange rates you hold the nominal exchange rate and foreign price level constant, and allow the nominal money supply to adjust endogenously, so it's the real exchange rate which changes. The reason for the downward-sloping AD in this case is exactly analogous to the downward-sloping demand curve for apples.)

Your answer is to hold the (expected) future price level constant (and presumably hold the nominal interest rate constant too), so the real interest rate is changing by definition. Interesting, but it's not the "orthodox" answer."

No, I said nothing about the expected future price level. My argument is identical to yours holding the money supply constant, this causes the nominal interest rate to fall and since prices are sticky a fall in the nominal rate is a fall in the real rate. It all works through the real rate and that is the standard textbook answer.

PS: there is no contradiciton in drawing the curve with P on the vertical axis and saying prices are sticky. The price level P enters individual maximization problems as a parameter. You choose a value of P, solve for the implied AD allowing interest rates to vary. The curve is traced out by repeating the exercise for each different value of P.

And Nick, the quote (of Krugman) that you yourself cite is not consistent with how you interpret Krugman in the post.

Krugman: "But if everyone takes a pay cut, nobody gains a competitive advantage. So there’s no benefit to the economy from lower wages. Meanwhile, the fall in wages can worsen the economy’s problems on other fronts."

He's clearly saying that wages and prices falling together has no benefit, he does NOT say it lowers AD, he says it changes nothing. Clearly he has in mind an unchanged real wage. The problems are "on other fronts". He then goes on to explain the problems on other fronts as the debt deflation problem.

Just to be extra clear I'll rephrase from my 7:05:

My argument is identical to yours holding the NOMINAL money supply constant, so the REAL money supply changes, this causes the nominal interest rate to fall and since prices are sticky a fall in the nominal rate is a fall in the real rate. It all works through the real rate and that is the standard textbook answer.

Adam @ 6.50: I totally agree with what you say here. But I would say that Bryan Caplan gets the theory of AD wrong, rather than saying it's not about AD.

@7.05: OK. I totally misunderstood you. Yes, in the standard ISLM derivation of the AD curve, a fall in P increases M/P which reduces real interest rate (holding expected inflation constant), which increases I and reduces S (moving along the IS curve). Aha! I have just now read your 7.18, and see we are saying exactly the same thing. Good!

@7.09: My interpretation of PK is that he starts out with a vertical AD (wage and price cuts do nothing), then complicates the model, to bring in debt-deflation, which modifies the AD curve so it slopes the wrong way.

"And the rising burden of debt will put downward pressure on consumer spending, keeping the economy depressed."

That sounds to me like he's saying that a fall in the price level (holding real wages constant, for simplicity) will reduce real AD via the burden of debt. (He's implicitly assuming either debtors have a higher mpc than creditors, or that the US has net foreign debt, which of course it does.)

Just to be clear, when I say the AD curve slopes the "wrong" way, I don't mean Paul Krugman gets it wrong; I just mean the AD curve is upward-sloping.

And sorry for misunderstanding what you were saying about real interest rates.

Well yes, my point was that Krugman's saying that the decrease in AD happens via the debt deflation channel and not from the usual AS-AD diagram.

Basically I first read your post to the end, then read the Caplan one you linked to, got annoyed and started ranting on your blog! Didn't mean to put Caplan's mistake on you.

"Caplan one you linked to, got annoyed and started ranting on your blog!"

Don't worry, a dose of Caplan does that to the best of us:

"The comparative political analysis was likewise lacking. Chairman of Cato's Board of Directors William Niskanen asked why different Western countries have significantly different economic policies. "I believe that levels of rationality vary from country to country," Caplan replied. The French, in particular, fare poorly in this measurement."

The Myth of Bryan Caplan's Seriousness
http://www.prospect.org/cs/articles?article=the_myth_of_bryan_caplans_seriousness

I take it that Caplan has never heard of Descartes. That, or he is just an intellectually dishonest bigot. either way...

Nick, I agree with the way you approach this issue. It all boils down to what you assume about monetary policy. Adam is right that there are no logical errors with Krugman's argument. But the argument is totally implausible for a gold standard regime, as you say, and almost as implausible for a fiat regime.
Krugman has criticized those who argue that the New Deal prolonged the Depression, and spoke highly of a Gauti Eggertsson study that argued FDR's high wage policy may have helped spur the recovery. In an earlier post I looked at FDR's high wage policy during the Depression. The bottom line is that FDR tried a high wage policy 5 times, and each time was a spectacular failure. The worst failure (late July, 1933) occurred under fiat money. So much for upward sloping AD curves:

http://blogsandwikis.bentley.edu/themoneyillusion/?p=48

I believe that there is a Bernanke put on the price level. He won't allow significant deflation. In that case wage cuts are expansionary. At the same time I agree with Krugman that wage cuts are a sign that something is seriously wrong. As Earl Thompson has argued, wage cuts are a sign that money is too tight.

The story told about the inflation problems experienced during the 1970's blames, at least in part, the wage/price spiral on union's ability to demand higher wages. During the depths of the depression, I'm sure everyone would have loved a little bit of inflation. So it seems to me that pushing wages up is not an unreasonable policy measure to try to generate inflation - especially given the state of economic understanding in the 1930's.

So what did they do wrong? Could using wage controls to generate inflation have worked (it obviously didn't)? Was the problem that they did it in a one shot increase? Would it have been better to legislate that wages will go up 2% a year, every year until we say otherwise and by the way we aren't likely to say otherwise for a very long time?


Patrick, I think you are right that they were using the model people have used to explain the 1970s. But I find that model implausible. In the U.S. unions were weaker in the 1970s than the 1950s and early 1960s (when inflation was low.) In addition, most people believe Volcker's tight money policy ended high inflation (although I suppose Reagan's anti-union stance might have played a modest role.) But inflation also fell sharply in the 1980s in many countries that didn't have anti-union governments.
Again, your explanation of motive may be right, but I think people put too much weight on wages as a cause of inflation.

Here's my model of the 1970s and unions (perhaps unions were weak in the US, but not elsewhere, like the UK, Canada, Australia(?)).

The central bank targets unemployment, a particular level it sees as "full employment", call it u^. Associated with u^ is "full employment" real output y^. So the central bank does whatever it takes with money supply and the rate of interest to keep real AD at y^. So, given that target, the AD curve is vertical at y^.

There is a natural rate of unemployment u*, and an associated natural rate of output y*. The LRAS curve is vertical at y*.

An increase in union power, and individual unions' success in raising relative wages increase u* and reduces y*, shifting LRAS left.

If u*>u^, so y*

Eventually inflation gets too high, and the central bank stops targeting u.

But for an awful long time (and I remember the 1970's) people (including economists) said that monetary policy had nothing to do with inflation.

In a sense they were right. Because they saw monetary policy targeting "full employment" as the "normal", thing to do. So if full employment is less than the natural rate, blame the natural rate for being too high. Not monetary policy for being too loose.

Something got lost. Half way down it should read:

"If u*>u^, so y*

Damn! It keeps losing it.

Halfway down it should read:

"If u* is greater than u^, so y* is less than y^, the vertical AD curve will be to the right of the vertical LRAS curve. The result is ever-accelerating inflation."

I am not sure whether you guys are familiar with an article which deals with a topic related to the one you are talking about. Just in case you are not, I am making it available to you: http://www.econ.kuleuven.be/ew/academic/intecon/Degrauwe/PDG-papers/Work_in_progress_Presentations/Flow-Stock%20Deflations.pdf
Best,
Ale

Nick, That's also the model of inflation I teach my students (from Mishkin's text), the central bank targeting real GDP at a level above the natural rate. I was always taught that it wasn't enough for there to be union power, but rather that there had to be increasing union power. But if that is your argument, then I entirely agree, I do think strong unions could shift LRAS to the left. It's just something I haven't thought much about recently, as U.S. union membership peaked (in percentage terms) way back in the 1950s.
One other issue is the "illusion vs. reality" question. I'm sure that when Keynesian policymakers saw their best laid plans fail in the 1970s, and saw what looked like adverse AS shocks (but were partly the normal shift up in Phillips curves with higher expected inflation) there was a tendency to look for obvious scapegoats. And union contracts with big fat multi-year pay increases look especially anti-social when you are trying to bring inflation down without causing unemployment. So I'm sort of half agreeing with you and half wondering if they weren't partly a scapegoat for a still poorly understood natural rate model. Probably some of each in countries with strong union movements.

Scott: following on from your last comment:

What really changed with the monetarist counter-revolution, was what policy instrument was to be assigned to what objective.

Pre-Friedman, the view (especially in the UK) was:

Fiscal policy targets real AD and unemployment,
Monetary policy is in a supporting role, to control interest rates and/or the exchange rate to target the composition of AD between consumption/investment and between net exports and domestic absorption
Then what about inflation? OK, as an afterthought, "industrial policy" (code for controlling union and other monopoly power).

Post Friedman, the view was:

Monetary policy targets AD and inflation,
Fiscal policy in a supporting role to target the composition of demand
Then what about unemployment? OK, as an afterthought, "industrial policy"..

So what really changed was the view about what policy instruments should be assigned to what objectives. But the whole debate was couched in terms of what causes what.


Why do people so often think they know better than Paul Krugman? I think he is wrong sometimes, but in my experience, when you think he is wrong, he is mostly right.

reason, agreed 100%.

I would add that, for me, what's amazingly annoying is that people always want to criticize without boethering to properly understand what the man is saying.

Furthermore, regardless of what you think of him or his public persona, Krugman and most really good academics are completely dedicated to the truth. I do think they sometimes have trouble admitting mistakes in public but in a class (where everyone is just trying understand), when you correct a mistake or point out something they've missed every one of them will thank you and mean it.

I should clarify my meaning, when I say "in a class (where everyone is just trying understand)...", I include the teacher as someone just trying to understand.

Krugman envy / Krugman worship / Krugman yearning / Krugman denial / Krugman gotcha / Krugman obsession

pick one

But it also just comes with the territory. If you are a top economist, with the most widely read and influential blog, then:

1. It is socially optimal that his views are subject to more scrutiny and criticism. (It matters much more if he says something that might be wrong than if some unknown says something that might be wrong).

2. It is individually optimal for others to subject his views to more scrutiny and criticism. (If you note what might be a mistake in what he says you get more attention than if you note what might be a mistake in what some unknown says.)

So we are all guided by an invisible hand...

The comments to this entry are closed.

Search this site

  • Google

    WWW
    worthwhile.typepad.com
Blog powered by Typepad