The Tabarrok/Cowen AS/AD model
Those who have taught intro to macro have probably noticed that the AS/AD diagram is not well-suited to explaining modern recessions. During most post-war recessions the rate of inflation fell, but remained above zero. The only exception is the 1974 recession, where inflation actually increased due to supply shocks. But if you try to use the AS/AD diagram to show how an adverse demand shock could cause a recession, it isn’t easy to do, as the graph implies any negative AD shock should lead to deflation.
The AS/AD model was set up to explain recessions that occurred when the dollar price of gold was stable, and inflationary expectations were close to zero. The best examples were the 1920-21, 1929-33, and 1937-37 depressions, all of which saw significant deflation. But what can we do with the post war recessions, which typical see a mere slowdown in the rate of inflation? One solution is to use the Phillips Curve diagram, but it has another weakness—it can only illustrate demand shocks, not supply shocks.
Ideally we’d like to have an AS/AD diagram that has inflation, not the price level, on the vertical axis. And the new Tabarrok/Cowen text does just that. This is Figure 12.10 of their macro text.
Note that the AD curve should have a slope of exactly negative one. It should intersect both the vertical and horizontal axes at exactly the same point (5% in this case.) The 3% Solow growth curve represents the long run trend rate of real GDP growth. The SRAS curve is drawn as in any other AS/AD diagram, showing that the short run impact of slower NGDP growth is somewhat lower inflation, and much lower real GDP growth. That is also a fairly standard assumption.
However, because Tabarrok and Cowen are dealing with rates of change, not levels, it is possible that you could have a negative AD curve. So in principle when they wrote this text in 2007-08, they should have made a four quadrant diagram. However, they undoubtedly noticed that NGDP had not declined (year over year) since 1949, so they were on pretty safe ground assuming that the Fed would not be crazy enough to allow a decline in NGDP anytime soon. And of course a one quadrant diagram is easier for students. Should their revised addition add three more quadrants? I do think they need one such diagram to show the recent recession.
[It’s funny how often when you notice a pattern, it breaks down. Remember all the articles in 2005-06 about how real estate prices hadn’t declined nationwide since the 1930s? And then there is “Goodhart’s Law.”]
Although the recession officially began in December 2007, I have argued that the early stages were mostly supply-side problems due to housing and energy—what Arnold Kling calls recalculation. But in August 2008 the entire economy weakened dramatically, and lots of labor got reallocated into “leisure.” Unfortunately we don’t have monthly GDP data, and in addition our price indices underestimate the amount by which the rate of inflation declines during demand-side recessions. But let’s do the best with what we have. From 2008:2 to 2009:2 real GDP fell at a 3.8% rate, and the deflator rose 1.3%. If we use 2008:3 to 2009:2, then real GDP fell at a 4.2% rate and the deflator rose only 0.6%. Let’s take a simple average and assume that between about July 2008 and May 2009 real GDP fell at about a 4% rate and the GDP deflator rose at about a 1% rate. So NGDP fell at a rate of 3%.
OK, so what would the most severe part of the recession looks like using the Tabarrok/Cowen model? Here is one way of thinking about the issue: Let’s assume that for some strange reason the Fed decided to adopt a monetary policy that was so contractionary that NGDP fell at a 3% rate over 10 months. And let’s also assume that other than this demand shock, there was absolutely nothing wrong with the economy. There were no supply-side problems. There were no banking problems. So what would the Tabarrok/Cowen model predict? Look at their graph and mentally extend the SRAS to the left, into the next quadrant. Notice that it appears to come pretty close to the actual position of the economy (assumed to be negative 4% RGDP growth and 1% inflation.)
What sort of shock would produce that outcome? A sharp leftward shift in AD, from plus 5% to negative 3%. And of course that is what happened. Also note that there are a wide range of possible slopes of the SRAS that yield roughly this outcome. Indeed if at minus 4% RGDP growth the SRAS showed anywhere from 0% to 1.5% inflation, then if NGDP fell by 3% you’d still end up with negative 3% to negative 4.5% RGDP growth, in other words, a recession much like the one we just had. To summarize, the Tabarrok/Cowen model suggests that a negative 8% AD shock would produce the sort of recession we just experienced, regardless of whether there were any banking problems, or any needed reallocation in the economy.
I’m not sure that I am willing to go that far. I do think that the banking problems had some effect on output. But I also think that most of the banking problems were actually caused by the fall in NGDP. In past posts I have argued that most of the recession was caused by falling AD, probably at least 80% of the fall in output. But even if NGDP had been kept growing at 5%, we might have had a bit of stagflation.
You might notice that I am having a bit of fun with the implications of the Tabarrok/Cowen model, because Tyler Cowen himself isn’t convinced by my interpretation of the recession. In his view only about 1/3 of the recession is attributable to falling NGDP, the other two thirds is due to real problems in the economy, such as financial system turmoil and the need to reallocate labor out of certain sectors of the economy. In fairness to Tyler, I should make two points:
1. This is a book designed to teach students basic tools like AS/AD. Just because they teach these tools, doesn’t imply that the graph is intended to perfectly model a messy cyclical phenomenon. These undergraduate models need to oversimplify a very complex reality.
2. If Tyler’s 1/3 estimate is correct, then I believe he would model the real problems in the economy by (temporarily) shifting the Solow growth curve (or LRAS) to the left. Indeed to a point somewhat below zero percent. In that case the inflation rate would be a bit higher than I assumed, but still less than 2%. So in practice the stylized facts wouldn’t be all that different from what we actually observed. And when you add in all the other factors that make macro so messy, obviously the simple exercise I engaged in isn’t really able to discriminate between my hypothesis and Tyler’s.
Nevertheless, I think it is useful to see that fairly standard macro models predict that a sharp fall in NGDP will produce results pretty similar to what we have actually observed, even without a banking crisis. It tells me that it is quite likely that aside from whatever damage was done by the banking crisis, the fall in NGDP made it much worse. Again, NGDP falling roughly 8% below trend could easily depress RGDP growth at least 6% below trend. And that is a very big deal. In addition, the fall in NGDP also worsened the banking crisis itself. So it hurt the economy in two distinct ways.
Now let’s return to the Tabarrok/Cowen diagram and ask; “What should the Fed have done?” You may notice that they drew the point of macroeconomic equilibrium at precisely the point where there was enough AD to create 5% NGDP growth. Hmmm, where have we heard that suggestion before? And their graph suggests that if you let NGDP growth fall below 5% you will get a slowdown, or even a recession. So the Fed should set monetary policy at a level where 5% NGDP growth is expected. This had been their strategy for several decades. But last fall the Fed abandoned the policy, letting NGDP growth expectations fall well below 5%, indeed even below zero.
I get frustrated by all these news articles and heterodox economists claiming that mainstream macro can’t explain the current crisis. The Tabarrok/Cowen textbook provides a precise prediction of what would happen if the Fed let NGDP fall 3%. And what actually happened is almost exactly what Tabarrok and Cowen predicted would happen under that sort of monetary policy. So if even an intro to macro text can almost perfectly explain this crisis, then what is the big problem with modern macro?
To answer that question we need to take one step backward. We need to ask why NGDP fell 3%. And this is (I think) what most people mean by the failure of modern macro. Much of modern macro assumes that low interest rates mean “easy money,” so Fed policy should have prevented this fall in NGDP. Milton Friedman and I believe that low interest rates are usually an indicator of tight money. Maybe it’s time we re-thought the whole issue of how to define the stance of monetary policy, as the current methods don’t seem very useful. Do we really want to keep teaching our students that money was “tight” during the German hyperinflation?
I don’t like using any kind of interest rate as an indicator of Fed policy. But if you insist (as most economists seem to) why not at least use ex ante real interest rates? Then we can very easily explain the crisis this way:
1. Between July and November 2008 the Fed adopted an ultra-tight monetary policy.
2. Real interest rates on indexed 5 year T-bonds rose from 0.5% to 4.2%
3. The tight money caused NGDP growth to plunge from its usual 5% to negative 3%.
4. The AS/AD model predicts you’d get about 1% inflation and negative 4% real growth.
5. That’s what we got, therefore modern macro nicely explains the recent recession.
So tell me, what’s wrong with putting this story in our textbooks? And speaking of textbooks, recall that I teach all my undergrads the following maxims from Mishkin’s best-selling textbook:
1. It is dangerous always to associate the easing or the tightening of monetary policy with a fall or a rise in short term nominal interest rates.
2. Monetary policy can be highly effective in reviving a weak economy even if short term interest rates are already near zero.
A new post on Free Exchange suggests that “economists” haven’t absorbed these basic lessons from the principles of Money and Banking:
I am thankful for the stimulus package, but I believe (as is no doubt abundantly clear at this point) that economists were far too quick to assume that monetary policy was appropriately stimulative and out of ammo.
I hope he is wrong, but I have to admit that this might explain why so many economists think modern macro has failed. Perhaps they don’t consider Mishkin’s textbook to be part of modern macro. In fact, modern macro (especially Woodfordian macro) was never tried.
PS. In the past I have used Mankiw’s excellent intro textbook. Obviously I face a big decision. I’m lazy, and I really hate to change textbooks. On the other hand I like the Taborrak/Cowen diagram much better than Mankiw’s standard AS/AD diagram. On the other hand Mankiw has given my blog some very nice links. On the other hand Tyler has been even more supportive. Decisions, decisions.
HT: Alex Tabarrok and Dilip
Tags:
1. December 2009 at 19:05
Ch. 12 from the Tabarrok/Cowen text can be found at–
http://worthpublishers.com/CowenTabarrokMacro/macro/samples/CT_macro_ch12.pdf
and Ch. 12 appendix–
http://worthpublishers.com/CowenTabarrokMacro/macro/samples/Cowen1e_Appendix12.pdf
Modern Principles: Macroeconomics by Tabarrok/Cowen–
http://worthpublishers.com/CowenTabarrokMacro/macro/index.html
1. December 2009 at 19:09
Scott: A few weeks ago I gave a macro lecture. It was a one day 10 hour (with appropriate breaks) lecture to MBA students.
I felt very lucky to have a copy (preliminary edition – uncorrected proofs) of Cowen and Tabarrok´s new intro macro text.
Their dynamic version of the AS/AD model where AD is represented by growth in M and V instead of C+I+G+NX and with inflation and RGDP growth in the axis (instead of inflation and output gap) was much more “student friendly” especially in the discussions of the stylized facts of the crisis and also to show when (like in the first half of 2008) inflation targeting may cause trouble.
1. December 2009 at 19:30
So I guess we can expect this recession to last a lot longer then most believe! The job market is destroying a lot of people in my neighborhood and from your post…that will continue.
1. December 2009 at 20:23
So your step 1 is that the Fed adopted a tight monetary policy in July-November 2008. But surely there must be a prior cause for that?
The Fed didn’t raise interest rates during that period (on the contrary) so that wasn’t the source of the tightening. And I am fully prepared to believe that the low interest rates do not indicate loose policy.
But if 5% interest rates were appropriate in July, and 1% interest rates in November were too high, something must have changed in between. Money became tighter during that period, for some reason not originating in the Fed.
How would you say this happened? Was it a change in expectations? Did people reduce their estimates of NGDP growth and/or inflation? Or did the money multiplier fall during this period, perhaps as a result of post-Bear Stearns and pre-Lehman problems in the interbank market?
If such a change did take place, it makes perfect sense that the Fed’s policy was too tight for the changed circumstances. It is very believable that the Fed wouldn’t realise this in time, as they didn’t have (or use) the tools to spot these changes in the market, or didn’t believe things could change so fast.
But I still want to know how money became so tight. Do you know the answer?
1. December 2009 at 22:04
1. I don’t know why no one has talked about the changing of the rules for handling the reserve requirements as part of the monetary tightening. It caused a lot of banks to have to badly need to raise cash. I saw it mentioned once in the WSJ around the time it was happening and then not at all.
2. And yes, you can see the massive spike in PPI as a result of the Fed’s shoveling money out the door to try to deal with the bank problems. Then when they tighten up again you can see the absolutely huge drop in PPI in the middle of 2008.
3. You also say that recalculation effects were the cause of the early recession. Do you mean to say that recalculation triggered the recession?
4. Anyway, wrt interest rates, the change in the discount rate changes the money supply through refinancing effects and such, but I don’t see how having low interest would have an effect. If you and Friedman are correct
are correct about high interest rates signaling loose money. (I can think of lots of arguments that agree with you there.) What do you mean by interest rates? Do you mean the Treasury rate, the interbank lending rate, average mortgage rates, expected earnings rates from mutual/hedge funds?
5. Isn’t the AS/AD model in agreement with recalculation? It seems to be they are saying the same thing, except recalc describes it using the language of micro and AS/AD with the language of macro. It seems the AS/AD abstraction is more useful to explain quantitatively what is happening, (even if it it tautological) and the recalculation/malinvestment model explains qualitatively what is happening. In other words the modern macro theory gives us how to price monopoly currency and the Austrian theory gives us what our legal/business policy should look like.
1. December 2009 at 22:14
@Leigh
“But if 5% interest rates were appropriate in July, and 1% interest rates in November were too high, something must have changed in between. Money became tighter during that period, for some reason not originating in the Fed.”
“But I still want to know how money became so tight. Do you know the answer?”
I know this answer wasn’t addressed to me, but I am opinionated enough to answer anyway. At first I wasn’t sure I agreed with Dr. Sumner on the money being tight, but have since changed my mind. The interest rate wasn’t what caused the tight money, the discount rate had very little to do with it. What caused the tight money could have been mostly three things.
1. The fed paying interest on reserves.
2. The change in rules on bank reserves and how they had to be dealt with
3. Massive destruction of wealth/recalculation (Recalculation in assets would cause increased demand for currency).
1. December 2009 at 22:31
I taught a three week intro to macro course in August using the new Tabarrok/Cowen book. I’ve taught the course before, both in regular semesters and short courses. I found students comprehending the dynamic AS/AD diagram much better than the Keynesian TE stuff I used to teach. They brought up great questions and I was able to answer pretty much all of them with this model. With respect to Dr. Mankiw, I’d say the intellectual honesty on your part and the improved comprehension on the students’ part is well worth the costs of switching.
1. December 2009 at 23:15
I only took intro macro and micro as a student, so I don’t know much about economics. However, I have a faint suspicion that there is an over-reliance on these purely graphical approaches, which can really get one in trouble at times. This may be the case even with economists. Is there any truth to this?
There can be obvious disadvantages in using more explicit models, not the least of which are unwieldiness and using unjustifiable precision, but are there are significant advantages as well?
I personally find I have more success modeling complex systems explicitly, though I do sometimes refer to graphs in a pinch.
2. December 2009 at 00:08
I’m all in favor of better intro to economics texts. I started college as an econ major but changed to lit because I hated my intro to macro text at University of Texas so much. Then I changed to math — without even realizing that math was an even better major for continued study of economics (Bad advising: I only figured out years later I would have been in a good position to continue to study econ.). In retrospect, I wish I had stayed with econ, but I found those early classes so boring I never thought of looking back. Now all I can do is read econ blogs and feel like an outsider on a field I wish I were an insider on.
Goodhart’s Law sounds like an earlier version of Victor Niederhoffer’s principle of ever-changing-cycles.
http://en.wikipedia.org/wiki/Victor_Niederhoffer
Niederhoffer is one of those anti-EMH guys us amateur traders look to hope for.
2. December 2009 at 02:27
Scott,
Just a random thought: if people can only figure out the stance of monetary policy by the use of interest rates, then perhaps one way of measuring the stance is to use the difference between the real interest rate and the Wicksellian stabilising interest rate (or ‘natural’ rate).
This might be better than using the level of real interest rates per se, as real interest rates themselves are not always the best indicator of the stance of monetary policy.
Nick Rowe explains this a lot more eloquently:
http://worthwhile.typepad.com/worthwhile_canadian_initi/2009/11/why-current-ad-depends-on-expected-future-ad-scott-sumner-in-islm.html?cid=6a00d83451688169e20120a668ed8f970b
Alan
P.S. Sorry for not replying to your posts for a while; I’ve continued reading them, but been a bit tardy with the comments!
2. December 2009 at 03:06
Doc Merlin: Thanks – you are of course welcome to answer any question you like!
I agree that interest rate changes did not cause the tightness – they were mostly a response to tightness arising elsewhere. In Scott’s view, which I’m sympathetic to, they were an insufficient response.
I don’t think interest on reserves were responsible, as that policy did not happen until October 2008. I’m not sure what change in reserve rules you mean, so I’ll leave that one.
The recalculation argument seems the most likely of your three, but I’m not convinced that the recalculation process could start so quickly – surely these kind of relatively major shifts in the economy take a while to kick in.
It turns out Scott partly answered my question in the previous posting (good monetarism, bad monetarism): “In the US in 2008 velocity fell sharply, so did the money multiplier”.
I think this is correct – V and/or m fell, the Fed responded with interest rate cuts but these were not enough – therefore policy was still too tight for the new circumstances.
But now I want to know: why did velocity fall? I guess because of a change in expectations about future NGDP. Why did expectations change? I don’t think this is a recalculation story – not directly at least. Although NGDP expectations are partly a function of what people think about recalculation.
It is attractive to think that people were somehow under a mass delusion, a bubble which simply burst, an “emperor’s new clothes” story. But it also sounds too simplistic. I do feel there is a strong psychological component to these events but I think we need to understand it a bit better than just to say “expectations changed”.
2. December 2009 at 05:56
Thanks Richard,
I am sure they’ll be glad to hear that marcus.
Leigh, In earlier posts I talked argued that the profession is intellectually bankrupt on the subject on the stance of monetary policy. Economists talk confidently about “easy money” and “tight money” without having any coherent explanation of what the terms mean. Just to be clear, I do NOT favor using real interest rates as an indicator of monetary ease or tightness. But if economists insist on using real rates as an indicator, then the Fed should have increased the MB enough to prevent the rise in real rates. As an analogy, suppose you used nominal rates as an indicator. Suppose you thought the Fed “doing something” consisted of raising and lowering its fed funds target. Also suppose the fed funds rate rose from 2% to 8% during a period where the base was constant. Many economists would say the Fed adopted a “tight money” policy, even though they actually “did nothing” in terms of the actual instrument of policy–the base. As you know, the Fed doesn’t control market rates, they merely target them.
Here’s another analogy. Most economists agree with Friedman and Schwartz that monetary policy was too contractionary during the early 1930s. But what did the Fed do? They sharply cut rates and sharply raised the monetary base. So my take on recent Fed policy is in the broadly accepted tradition of F&S.
So here’s what the Fed should have done to avoid a tight money policy. They should have raised the base enough during July to November so that NGDP growth expectations stayed at around 5%, instead of falling into negative territory. I think NGDP growth expectations are a much better indicator than either real or nominal interest rates. But again, people are free to use whatever indicator they wish. I mere pointed out that if we are going to use nominal rates, then we must teach our students that money was tight during the German hyperinflation. And if we are going to use real rates, then we must teach our students that money was very tight during July-November 2008.
How did money get tight? The Fed let NGDP expectations fall sharply, when they had the tools to prevent that from happening. Call that an error of omission? Yes, but so is the Fed raising rates from 2% to 8% in a period when the base is constant. That is also an error of omission, it is not the Fed literally “doing something.”
Doc Merlin, I agree about the PPI, and lots of other market indicators showed the same tightness–the dollar soared against the euro, for instance.
Recalculation and the energy shock triggered a mild slowdown. (There was also a very mild demand shock in early 2008) Without the plunge after July 2008 there would have been no outright recession. In an earlier post I discussed how recessions are still preventable after they have begun, because a ‘recession’ requires a large and sustained period of weakness–and that only occurred after mid 2008.
By interest rates I mean Treasury yields.
No, recalculation is not saying the same thing. Kling says that even if the Fed had kept NGDP growth from falling, we would still have had a recession. My view is that we would not have had a recession if NGDP had risen at 5%.
Thanks Norman.
Mike, I think grpahs give a nice overview of how changes in NGDP are partitioned into changes in P and Y. Then we can get into the details of why NGDP changed, which is a much more complex issue.
rob, Thanks for the info.
Alan, Yes, if people insist on using interest rates, that is probably the best way. But I would hope we can look beyond interest rates. Most economists already know the theoretical problems with using nominal rates, but in a crisis they still fall back on that short cut, as we just saw. The spread you refer to is hard to measure, so I worry that people will take shortcuts.
Leigh#2, You said:
“I agree that interest rate changes did not cause the tightness – they were mostly a response to tightness arising elsewhere. In Scott’s view, which I’m sympathetic to, they were an insufficient response.
I don’t think interest on reserves were responsible, as that policy did not happen until October 2008. I’m not sure what change in reserve rules you mean, so I’ll leave that one.”
I agree that higher real interest rates were a response to tightness elsewhere, but I don’t necessarily agree that they we an insufficient response, as that presupposes the Fed should target interest rates. Suppose the more aggressive response had been the Fed increasing the base a great deal, or setting an explicit 5% NGDP growth target, level targeting. Those sorts of aggressive moves might have lowered the real rate, indeed probably would have. But it is theoretically possible that they would have so boosted bullish expectations that the real rate would have risen a bit. This is why I prefer to avoid interest rates. High real rates can indicate tight money, but can also indicate strong real growth expectations.
On the question of interest on reserves, I agree that the policy occurred too late to trigger the initial crash. But the announcement on October 3rd did roughly correlate with the stock market crash in the first ten days of October, so it might have played a supporting role.
2. December 2009 at 08:36
Hi Scott,
Ive been mulling over the whole idea of targeting NGDP rather than the inflation rate, which does make sense too me once I think about it; its somewhat of a new concept for me.
But, my question is….why not target real gdp rather than nominal? In fact, why not target real so its equal to nominal? Would this not ensure growth while also keeping inflation low?
Thanks,
Joe
2. December 2009 at 08:59
The bond market is now indicating that the longer-run AS curve is well to the left of the SRAS. And by “longer-run”, I mean out one year.
The reason is clear: we are at 2% inflation expectations on the 5yr TIPS, and the 10yr-3mo yield curve is as steep as it has been in the past twenty years according to the NY Fed model. So obviously, the bond market expects there to be little growth coincident with 2% inflation. Or, if you like, little growth coincident with 0% short term inflation and 3% longer term inflation. Either way, there is no evidence to support the notion that the bond market expects the slope of the 5-yr AS curve to be shallow. In other words, we are looking at a minimal response of AS to stimulus.
So, your response is to raise short-term inflation expectations, and therefore capitalize on the shallow portion of the SRAS (gaining growth with little inflation impact). What effect would this have on longer term inflation expectations? Obviously, to raise short run expectations to 2% would involve risking even higher inflation over five years. Given that the longer-run AS curve is ALREADY seen as quite steep, the result could be much higher (4-5%) longer term inflation expectations, combined with 2% short-run expectations. This would correspond with a yield on 10yr bonds of close to 5% assuming a required real return of at least 1%.
So, at the end of the day, what you, Paul Krugman, and other Central Planners have in common is the desire to:
-reduce real incomes by combining inflation with sticky wages
-reduce real disposable incomes by increasing nominal debt service (in the form of higher nominal rates)
I can see how macroeconomics would have, as its root “market clearing” solution to unemployment, the screwing of consumers. I can also see how, as a group, economists could be so unbelievably short-sided as to think an over-levered consumer can possibly withstand such a blow. Maybe your guess is that employment/wage growth in manufacturing exports — a smidgen of the economy — can make up for a continuing consumer spending recession.
2. December 2009 at 09:12
I guess I should have written my macro book.
I’ve argued for years that the axes of the AS/AD diagram should be rate of inflation and rate of growth in GDP. Taylor’s book gets it half-right (rate of inflation), but continues to have level of real (I think; I can’t find my copy of the book right now) GDP on the horizontal axis. I may have to consider making the change as well…
Decisions, decisions…
2. December 2009 at 13:06
“No, recalculation is not saying the same thing. Kling says that even if the Fed had kept NGDP growth from falling, we would still have had a recession. My view is that we would not have had a recession if NGDP had risen at 5%.”
My view is we still would have, due to unemployment increase and the housing bubble collapsing, triggering a wave of bad bets/debts being called in. The housing market was in huge trouble, and so was student loan repayment. I would also argue that the spike in early_2008/late_2007 made the plunge inevitable due to Hayekian trade cycle and production posibilities frontier type arguments.
Anyway, I think all of this is secondary to fears over expected legal policy. If you want we can make a bet. I bet that as soon as it looks like the legal policy initiatives the democrats are trying to pass all look dead in the water, the economy will take off like a rocket, particularly in the more free market states.
2. December 2009 at 13:40
I don’t understand the SRAS curve in that model.
The normal Phillips curve has
inflation = expected inflation + b(y-y*)
This one has:
inflation = expected inflation + b(ydot – y*dot)
3. December 2009 at 07:09
Joe, You cannot target Real GDP because if you did the price level would go to either zero or infinity. Here I am assuming that the estimated “natural” rate of RGDP is at least a tiny bit different from the true natural rate of RGDP. I think my views here are now accepted by all macroeconomists, so what I am saying is no longer controversial (although it might have been in the 1960s.)
David, You said;
“The reason is clear: we are at 2% inflation expectations on the 5yr TIPS, and the 10yr-3mo yield curve is as steep as it has been in the past twenty years according to the NY Fed model. So obviously, the bond market expects there to be little growth coincident with 2% inflation. Or, if you like, little growth coincident with 0% short term inflation and 3% longer term inflation.”
You lost me somewhere. Doesn’t a steep yield curve usually signal strong growth, not weak? FWIW, I think the TIPS are forecasting 1% inflation for two years, then 2.3% inflation.
you said;
“So, your response is to raise short-term inflation expectations, and therefore capitalize on the shallow portion of the SRAS (gaining growth with little inflation impact). What effect would this have on longer term inflation expectations? Obviously, to raise short run expectations to 2% would involve risking even higher inflation over five years. Given that the longer-run AS curve is ALREADY seen as quite steep, the result could be much higher (4-5%) longer term inflation expectations, combined with 2% short-run expectations. This would correspond with a yield on 10yr bonds of close to 5% assuming a required real return of at least 1%.”
My number one suggestion is level targeting. Otherwise we swing from too little inflation to too much inflation (as you said.)
But I think we are still a long way from too much inflation. A watched pot never boils. I think the Fed is so overly obsessed with inflation right now that they will snuff out any gradually rise in inflation above 3%.
You said;
“So, at the end of the day, what you, Paul Krugman, and other Central Planners have in common is the desire to:
-reduce real incomes by combining inflation with sticky wages”
When FDR tried this policy in early 1933, real hourly wages fell sharply (until the NIRA kicked in), but total real income rose sharply. That’s what we need now. Deflation has made real wages too high. Inflation (better yet NGDP growth) can undo that damage.
I don’t understand your last paragraph at all. Those are certainly not my views. If we get growth, consumer spending will probably rise.
Donald, I had favored actual minus expected NGDP on the vertical axis, and hours worked on the nominal axis. This is because I favor a macroeconomics free of concepts like inflation and interest rates.
Doc Merlin, In earlier posts I presented evidence that the fall in NGDP caused the biggest part of the housing crash. And it certainly caused the big jump in unemployment. Unemployment did not rise sharply after the recession began in December 2007. Rather it rose sharply after NGDP started falling in August 2008.
Nick, I don’t work with that sort of Phillips Curve model, as I think it confuses cause and effect. (Output doesn’t cause inflation, inflation causes output.) I prefer the Milton Friedman version. But I would point out that when you have a clearly identified monetary shock then the second equation seems a much better fit. Example; dollar devaluation in 1933:
1 Output 30% below capacity.
2. Growth rapid
3. Inflation rapid.
Which one explains those stylized facts?
3. December 2009 at 09:53
Scott: “I think the Fed is so overly obsessed with inflation right now that they will snuff out any gradually rise in inflation above 3%.”
Unless Bernanke is lying, the Fed is not concerned about inflation. He doesn’t see any signs of inflation in the near future. That’s what they say, anyway.
Scott: “The Fed let NGDP expectations fall sharply, when they had the tools to prevent that from happening.”
Why you think Bernanke failed/refused to respond to the the drop in inflation expectations as you accuse him? He could see the same “shocks” that you saw.
3. December 2009 at 10:18
For what it’s worth, Chad Jones’ intermediate macro textbook
http://books.wwnorton.com/books/detail.aspx?ID=13800
also has an AS-AD diagram with inflation on the vertical axis, so it’s not as if this is the first time this variation has made it into a textbook. (And Chad’s book isn’t written too far above the level of an introductory text.) I teach using that book but have no other connection to it, so this isn’t an ad.
3. December 2009 at 11:24
Scott: I would say that Friedman was strategically ambiguous. Was he talking about a Phelpsian disequilibrium price adjustment process, or a Lucasian supply curve? In either case, you get the same equation. It’s my equation 1, just rearranged so y-y* appears on the left hand side.
But your empirical counter-example is important. It’s a counterexample to both Phelps and Lucas. It does fit your sticky wage story better, I think. What happened to nominal wages over the same period?
3. December 2009 at 13:39
fundamentalist, You misunderstood Bernanke. He didn’t say that he doesn’t care about inflation, he said that he doesn’ty expect it. Neither do I. If inflation expectations rose to 3% in the TIPS market they’d raise rates until inflation expectations fell back below 3%.
Obviously the Fed didn’t respond aggressively to the collapse last fall. They refused to cut rates at all at their September meeting, even after Lehman had failed. If they had wanted to prevent NGDP growth falling below 5% surely they would have cut rates at that meeting. By early October they started paying interest on reserves, which was a contractioanry policy. The target rate was still 2%, when it should have been zero.
William, Thanks, but does he also have the growth rate for real GDP on the horizontal axis? That is important.
Nick, You said;
“Scott: I would say that Friedman was strategically ambiguous. Was he talking about a Phelpsian disequilibrium price adjustment process, or a Lucasian supply curve? In either case, you get the same equation. It’s my equation 1, just rearranged so y-y* appears on the left hand side.
But your empirical counter-example is important. It’s a counterexample to both Phelps and Lucas. It does fit your sticky wage story better, I think. What happened to nominal wages over the same period?”
Yeah, you are right. But would you agree that Friedman’s view of causality naturally leads to versions of the Phillips curve where you can get rapid inflation during high unemployment. If money leads to inflation directly (say through currency depreciation) and if RGDP changes slowly, then the stylized facts will fit the M -> P -> Y mechanism.
Regarding your second point. Unfortunately, it is a very short period, but the results were so striking that I still think they are meaningful. Nominal wages were flat from March to July 1933, the WPI rose 14% and industrial production rose 57%. Then the NIRA ruined everything, otherwise the Depression would have ended quickly.
BTW, Rapid dollar devaluation drove the whole process.
3. December 2009 at 14:43
Scott: “If they had wanted to prevent NGDP growth falling below 5% surely they would have cut rates at that meeting.”
So why do you think the Fed did not want to prevent NGDP from falling?
3. December 2009 at 22:34
Scott, long-time reader, first time commenter. How do you respond to the criticism, raised both in the comments on MR and somewhere else I can’t remember, that the problem with Sumnerianism is Goodhart’s Law? If the Fed targets NGDP, then NGDP will lose its significance as the relevant indicator–or something like that.
I admit I don’t really understand Goodhart’s Law, but I think the other critics might. Since you bring it up in this post, perhaps you should offer a short explanation why it wouldn’t apply to NGDP targeting.
5. December 2009 at 11:41
fundamentalist. In a way they do want to prevent NGDP from falling. But they aren’t willing to take aggressive steps to make it happen. I suppose there are any number of reasons for their caution. See my new post on Wall Street.
rob, Goodhart’s law doesn’t apply to NGDP, because it is (by assumption) the goal of policy to stabilize expected future NGDP along a 5% growth trajectory. Goodhart’s Law applies to cases where you target an intermediate variable in the hope that it will continue to be correlated with the goal variable even after you begin targeting it.
Indeed Goodhart’s Law is one of the best argument’s for NGDP targeting.
12. August 2010 at 08:03
[…] weakens, but the central bank decides not to respond. That is the issue now.)Scott Sumner would quibble with the assessment that the Fed was unable to prevent the 2008 collapse, and I don't know that […]
5. September 2012 at 04:00
[…] enjoyed this post by Scott Sumner, and wanted to use it as an application. It is clear to me how to show a purely nominal explanation […]
9. April 2013 at 15:52
[…] Alex Tabarrok reminded me that the Tabarrok/Cowen text is an […]