QE2 and the Laffer Curve.
I am not able to get anyone to debate me on monetary policy in a liquidity trap. Therefore I resort to crude provocation.
I recall two claims about monetary policy which were not controversial until this year. First that the effects of a shift in monetary policy peak after roughly 6 months. Second that it acts through investment and especially housing investment (the second follows from the view that it acts via interest rates but not the short term rates which the Fed can control but medium and long term rates which matter a lot for housing and some for investment in productive capital).
After fiddling with the dates to avoid the inconvenient fact that medium term nominal interest rates went up when the actual QE2 purchases began the money supply side economists decided that the date the policy began was late August. That means that the data you present here would, in a sane world, be the last nail in the coffin of the hypothesis that the Fed can stimulate the US economy when it is in a liquidity trap by buying 7 year Treasury notes.
I get rude after the jump.
I note already that 7 year real interest rates are higher than they were when the actual purchases started and equal to what they were when Bernanke first mentioned QE2 (late August).
So I ask you what evidence could possibly convince you that QE2 was ineffective. You have a powerful imagination and should be able to think of something (I can’t).
I also note that the quality (intelligence and integrity) of data analysis of the money supply side economists reminds me of the data analysis by supply side economists. I think they have resorted to each and every cheap rhetorical trick used by the supply siders. For every bit of bogus supply side data abuse, I claim I can find the same bogus trick played by advocates of quantitative easing.
If I am presented with no such examples, I will conclude that the money supply siders admit that they as bad as supply siders. I will present my examples of similarly bogus data analysis tomorrow.
update: as promised, they are posted. But since they are not up to angrybear levels of seriosness and civility (would you believe raging Shrew ?) they are posted at Robert’s Stochastic Thoughts
Why the assumption that an anaklysis of any set of data, raw or central tendencied, must be valid? Do economists that publish in professional hopefully scholarly journas) ever see their work subjected to peer review. Or are they simply devided into theoretical (should read that as ideological) camps? If one starts from an ideological position the data can be interpreted to mean what ever one’s ideology would want it to mean. Peer review isn’t an operative funtion when big money may be in play.
well this is delicate. I just received a rejection letter from a peer reviewed journal (what’s worse a summary rejection by an editor who didn’t even bother referees). There definitely are rules describing what economists may and may not do with data. I have no clue what they might be (hence the shortness of my CV).
There are, as you note, different schools. They differ not only in their favored theory (or ideology) but also in the views on methodology. Some are frankly theorists who consider economics (or at least their field of economics) to be a branch of mathematics. They don’t claim that theiretical work corresponds in any particular way to mere reality.
Some strongly believe in the no one true correct specification view. The approach is to start with very simple summary statistics (such as simple correlations) and work up to multiple regression and then instrumental variables. Only results which look similar at all stages of sophistication are treated as worth attention. These people are, I swear, actually able to learn things from data. However, they don’t deal with macroeconomics where there are too few data for there to be results which are robust in this sense (some economic historians use this approach with very old macro data — they all seem to work in Berkeley IIRC).
Macro economists are sharply divided into three methodological camps (four counting the few pure theorists — I have written purely theoretical macro with absolutely no claim that it has any real world relevance and that was published). There are the “structural” macroeconometricians where “structural” means “contaminated with theory.” There are atheoretical time series analysts who are all very irritated that Christopher Sims managed to get famous by naming a regression (he called it a VAR and they give a name to every summary statistic they calculate). They have shown that with no theory and few data one can reach any conclusion one wants. Finally there are the simulators who are practically theorists. They have computers approximately solve extremely complicated models and then calcuate a few summary statistics of the simulated data and compare them to summary statistics calculated from real data.
Each of these groups thinks the research of the other two groups is worthless. So they are right about something.
The problem with the real world is that you only get to run one experiment at a time. It is frustrating, in this case, that I don’t have a clue what would have happened if nothing or something else were done.
This is a straw man argument.
“I recall two claims about monetary policy which were not controversial until this year. First that the effects of a shift in monetary policy peak after roughly 6 months.
Most (well over half) of published macro is by New Keynesians these days. The principle of rational expectations is accepted by all New Keynesians. If one accepts the principle of ratex one expects policy changes to start impacting the economy as soon as they are rationally expected. If you do not accept the principle of ratex than say so. But please remind your readers that that puts your squarely in the minority of the macroeconomics community.
Yes, empirical analysis on the impacts of monetary policy in the postwar period suggests there is probably about a 6 month lag between policy implementation and peak effect. This fact does not contradict the other.
“Second that it acts through investment and especially housing investment (the second follows from the view that it acts via interest rates but not the short term rates which the Fed can control but medium and long term rates which matter a lot for housing and some for investment in productive capital).”
The interest rate channel is not the only channel of the monetary transmission mechanism. There are other channels, such as the inflation expectations channel, the asset price channel the exchange rate channel etc. There is nothing controversial about this, as every undergraduate intermediate textbook on monetary policy that I’m familiar with, mentions these channels. These channels haven’t suddenly been whipped up in the past few months. This is standard monetary economics that has been around for decades.
“After fiddling with the dates to avoid the inconvenient fact that medium term nominal interest rates went up when the actual QE2 purchases began the money supply side economists decided that the date the policy began was late August”
I know of not one single Quasi-Monetarist (Sumner, Beckworth, Woolsey, Kantoos, Nunes etc. etc.) who thought QE2 would bring down medium term nominal interest rates ex ante. Every single one believed if QE2 raised nominal GDP expectations that this should raise medium term nominal rates (duh). I am aware there are those who think that the purpose of QE2 was to lower medium term nominal interest rates (e.g. Hamilton). But those are people who put heavy emphasis on the interest rate channel of the monetary transmission mechanism. Clearly, Quasi-Monetarists are not among those.
If you want a debate with Quasi-Monetarists perhaps you sould do them the favor of first properly representing their claims as well as what is in fact the consensus in monetary economics concerning the monetary transmission mechanism. Implying that your views are themselves mainstream is a joke.
Mark my claim should have been that the effect of a shift on monetary policy on GDP peaks at 6 months. I think the reference to “investment” makes it clear that I was discussing NIPA flows not asset prices. Still I think I should have included the words “on GDP.”
First I cite Milton Friedman who is considered by many to have had a bit to say about monetary policy “long and variable lags.”
Second I quote Paul Krugman who is considered by me to have a bit to say about monetary policy http://bit.ly/kJFhYW
Third Michael Woodford (I have not begun to search — I name him as a leading new Keynesian). http://www.nber.org/papers/t0233
I have met only one economist who ever claimed to believe in rational expectations (and I think he was joking). All other economists who have discussed the matter assert that the assumption is not literally true. Many go on to say it is useful, but no one else has claimed that it is true. In particular Thomas Sargent was very very clear on that point the last time we had lunch (OK it was also the first time we had lunch but I think you will agree that he is not a completely minor figure in the rational expectations school).
I do not believe in rational expectations or the tooth fairy. I believe I have made that clear many times. However, it is also true that new Keynesians fiddle their models so that the effect of a shift in monetary policy on GDP peaks after 6 months, because that is a stylized fact. Like all facts, it can be reconciled with the rational expectations assumption if one makes the right assumptions.
Your first comment demonstrates astonishing ignorance of developments in Macroeconomics in the past 50 years. It should be clear what an macroeconomist means when he says “effects of monetary policy peak at 6 months.” I did leave a word and an acronym out, but the stylized fact is so well known that anyone who is not amazingly ignorant about macroeconomics would have known that the claim is an accepted stylized fact and so one which people make sure their models fit.
*that is the so called hypothesis is nothing of the kind. When I cited one Noble laureate on this point (Sen) my advisor (Katz) snorted and said that Solow told entering MIT graduate students that as one of the basic things they must know from the very start).
I guess I just don’t get it, but I do not understand what QE II was supposed to do for the Main Street economy.
Robert – help this dummy.
Obviously I was talking about the effect of monetary policy on GDP. And yes I knew by “investment” you were referring to physical investment and not financial investment. Being aware there is an asset price channel to monetary policy doesn’t mean one isn’t ultimately talking about the effect of monetary policy on output and incomes.
“I have met only one economist who ever claimed to believe in rational expectations (and I think he was joking). All other economists who have discussed the matter assert that the assumption is not literally true. Many go on to say it is useful, but no one else has claimed that it is true. In particular Thomas Sargent was very very clear on that point the last time we had lunch (OK it was also the first time we had lunch but I think you will agree that he is not a completely minor figure in the rational expectations school).”
This suggests that we live on different planets. I can honestly say I have talked to very few credentialed macroeconomists who do not believe in ratex in at least some form. Also I sincerely doubt Thomas Sargent said what you claim. (I suspect you heard what you wanted to hear.) Otherwise this would be pretty big news.
Your misinterpretation of my first comment is stunning and the conclusion you based on it is completely ill founded. And, I might add, this seems to be a consistent pattern with you. (I wonder if it is willful, or merely due to an exceptionally poor ability to make accurate inferences from verbal communication.)
Also, you can quit with all the ridiculous condescension and the silly name dropping. It might be useful for you to remind yourself you’re not the only person in the universe with a PhD in macro from a reputable institution.
Me too! I’m still trying to figure out what the intent of QEII was in general; never mind the effect on main street. I suspect that Bernanke figured that if he gave boatloads of money to Wall Street, main street would prosper? Here’s a discussion from Dec at David Beckworth’s blog on QE II in which I argued the intent was to lower longterm rates thru inflation in the comments. David Woolsey’s response to my comment has me completly and utterly confused.
“QE2 and Rising Yields”
I jump at the crude provocation.
The liquidity trap is just a piece of nonsense. The logic behind the speculative motive of liquidity preference, from which the liquidity trap derives, applies only to a system in which currency is the sole form of money. When banks and demand deposits are introduced into the system, the logic behind liquidity preference collapses. I have explained at length at http://www.philipji.com/item/2011-04-17/paul-krugman-and-the-liquidity-trap
Ever since the Great Depression bank rescues have played a key role in preventing recessions. I wonder whether macroeconomists have ever paused to ask themselves why in all of the General Theory, John Maynard Keynes makes no mention at all of bank rescues, whereas Bagehot writing more than half a century earlier gave the subject the utmost importance.
The answer simply is that Keynes’s model is a crude currency model in which banks don’t exist. That is why although there is a lot of space given to real wages and monetary wages, to which no modern economist pays any attention, there is no mention of the importance of saving banks.
There’s a simple way to test Keynes. The next time there is a financial crisis let the banks and financial institutions go under and just run up a large fiscal deficit. Of course this will require great courage and great stupidity.
I don’t think you are alone at all in not understanding QEII. I think the most interesting thing about QEII is the vast diversity of views about what it was and what it was intended to accomplish. IMHO, the difference between you and most folks opining on QEII on the internet is that you know you don’t understand it whereas most everyone else thinks that they do understand it. It seems unlikely that more than one or two of them are correct as they all seem to see something different.
The St Louis Fed had a report on the effects of QE lat week. I thought this line from the report summed things up very well:
A less-recognized risk in LSAP programs is that permanent increases in the monetary base foreshadow eventual increases in inflation that can increase, rather than reduce, unemployment over the long term.
In other words, QE has a negative long-term impact on unemployment. Put that in you pipe Mr. B and smoke it.
The quote you’re referring to is here:
“A less-recognized risk in LSAP programs is that permanent
increases in the monetary base foreshadow eventual
increases in inflation that can increase, rather than reduce,
unemployment over the long term. David Ranson of
Wainwright Economics has analyzed the U.S. data over the
period of 1950 through 2007. Ranson divided the 57-year
period into two categories: years when the monetary base
grew at an above-average rate (8.1 percent) and years when
it grew at a below-average rate (3.5 percent). Ironically,
economic growth was higher in the years of slow money
growth (3.7 percent) than it was in the years of rapid growth
(3.2 percent). The same was true for industrial production.
Meanwhile, the consumer price index rose 5.1 percent in
years of above-average monetary growth and just 2.6 percent
in below-average years. The gold price showed an even
bigger differential, rising 12.5 percent in above-average
years and just 0.6 percent in below-average years.”
First, Yi Wen draws this conclusion based on David Ranson’s analysis. But David Ranson’s analysis does not show that *permanent* increases in the monetary base lead to lower growth. Moreover, advocates of greater monetary stimulus are not talking about targeting the monetary base. Most are talking either about price level targeting or nominal GDP level targeting.
Second, there is reason to be sceptical about this is the source: David Ranson. David Ranson is President of Wainwright Economics, a brokerage with an Austrian bent (they employ John Tamny) that advocates a return to the gold standard. Although Ranson has a PhD in business economics he has little formal background in macroeconomics. His analysis is private, unpublished, and not peer reviewed.
Third, the monetary base doesn’t move off trend that much, and when it does its typically due to countercyclical Fed action:
The base drops as the Fed tries to cool an overheating economy, rises as the Fed tries to perk up a lagging economy, and soars when crisis strikes. It should be obvious that growth is generally the response to, rather than the cause, of an expectation of slowing growth.
Yi Wen goes on to cite his own research, arguing that inflation causes unemployment, but offers no credible reason of his own why this is so. But he does site a recently published paper by Berentsen, Menzio and Wright called “Inflation and Unemployment in the Long Run.” (American Economic Review, February 2011, 101(1), pp. 371-98). These authors use a search-and-matching model to explain why longer-term inflation can increase, the unemployment rate. That is, inflation reduces the demand for money and, hence, hinders trade and the probability of matches in both the goods and labor markets. They account for a sizable fraction of the increase in unemployment rates during the 1970s, but fail to find significant evidence for this effect in other decades.
As usual context is key. How is a model that found evidence for this effect during a time period when inflation was historically high and […]
“So I ask you what evidence could possibly convince you that QE2 was ineffective. You have a powerful imagination and should be able to think of something (I can’t).”
I turn the microphone over to Brad Delong:
“If market expectations of future inflation had not risen but had declined since QEII was mooted last August, I would conclude that QEII has been ineffective. They have not.”
Well it made some sense (except for the part about buying the assets closest to T-bills which had a positive interest rate). The idea is print Fed deposits (electronic money) and use it to buy assets with positive interest rates. This should drive down those positive interest rates (didn’t happen) and cause higher exepected inflation (either as a result of lower interest rates and etc and a Phillips curve or because of irrational respect for the ghost of Milton Friedman). The two effects sum to lower real interest rates.
Low real interest rates make it better to invest. Especially to invest in a house (as in buy one). . Also and less so for firms to build new factories or stores or offices or whatever.
Higher investment means higher aggregate demand which means higher GDP which helps main street. Especially higher demand for houses (which makes sense if long term real interest rates are low) makes the houses and appartments on main street worth more, which is great for the poor suckers who bought them and have to refinance or default.
It all makes some sense. For example QE1 worked wonderfully helping to prevent Great Depression II (this time it’s your personal house). and making roughly 500,000 killings for the US federal government (I’m not in the market and I don’t know how much one earns for a hit, but I think you can get someone killed for about $100,000).
QE1 worked millions of killings. QE2 was a flop. The difference is that the Fed has to buy something which scares private investors (so they don’t want to hold it) not the very safest asset which pays more than zero return.
They desicded to focus on quantity ($600,000,000,000 is a lot) not quality (7 year Treeasury notes are extremely high quality so no one cares who owns them). If they had bought risky garbage with QE2 as they did with QE1 we would have robust growth and a pony.
thanks for the jump. As far as I can remember at the moment, Keynes didn’t write much of anything about banks, bank rescues, the money multiplier or the FDIC in the General Theory. In his defence, I might note that Darwin didn’t anticipate Mendel either. Keynes was very very very British. The particular problem of bank failures couldn’t interest him, because it occured only in the barbarian rest of the world. Bagehot wrote back just when the Bank of England was figuring out what to do. Keynes considered a pre-BAgehot analysis irrelevant to all important countries (the UK, England and the British Empire).
However, there is, in fact, a liquidity trap. Nominal interest rates can’t fall below zero. It works fine in a world with banks and, indeed, in a world without money at all (the really new new Keynesians don’t have money in their models — just a nominal interest rate — money is just a unit of account not a means of exchange which is achieved via barter at nominal rigid prices because well hell I can’t justify them I just describe them (they being mainly Woodford)).
Keynes did not stress the liquidity trap. He mentioned it twice, once in an appendix on Professor Marshall’s theory of the interest rate. The true apostle of the liquidity trap is Paul Krugman. He does not use crude currency models. He describes what happened in Japan in the 90s and in the USA now. Don’t pick on Keynes. Try to refute Krugman. It may be possible (as pigs may learn to fly for all I know).
The speculative demand for money as described by Keynes is not at all the same thing as the liquidity trap. He, unlike his critics, had actual experience as a (very succesfull) speculator. It has nothing to do with holding money when the highest available nominal interest rate is zero (that is when pigs fly). It is holding liquid assets while awaiting new information to avoid paying broker’s fees or bid ask spreads. This is obvious to everyone who has ever managed a portfolio and incomprehensible to the reprehensible idiots who think they can understand what Keynes wrote about speculation without speculating.
Keynes kept some of his assets and some of the assets he managed for Cambridge in money so he could react to new information without paying brokers the huge fees they required back when they had a cartel and succesfull conspiracy to restrain competition. He noted that he had done so and speculated that he wasn’t the only one. This has nothing at all to do with the liquidity trap. It is an undeniably accurate description of money holdings by speculators. One can deny it as one can assert that the Earth is flat. But both sides don’t have a point.
I have replied to Brad at his site (click the link) and my personal blog. QE2 was supposed to drive two asset prices up. One fell and the other (to be as charitable as possible on the timing of QE2) stayed about the same.
If one looks only at the difference one finds the way to process data to make QE2 look good. But there is no justification for processing the data in this way.
If I were to go to your house and piss mostly in your toilet and half on your floor and argue that the piss in the toilet minus the piss on the floor has increased so you should thank me, you would be pissed.
“QE2 was supposed to drive two asset prices up.”
I prefer to think about what it was likely to do. I think it’s clear by now that it didn’t do exactly what the architects expected.
From a Quasi-Monetarist perspective QE2 did the following:
1) Asset prices
Stock prices are up about 30% since Jackson Hole, adding an estimated $5 trillion to household wealth.
2) Exchange Rate
The trade weighted value of the dollar has fallen about 6% since Jackson Hole, making exports slightly more competitive. And in fact exports are up at a faster rate than imports the last two quarters, although the trade imbalance has actually widened owing to the fact that exports are starting from a lower base.
3) Inflation Expectations
Two year inflation expectations have risen from 0.9% before Jackson Hole to 2.7% today according to the zero coupon inflation swap. This is important because it eases the debt deflation problem.
4) Nominal GDP expectations
Unfortunately we have no market based measurement for NGDP. However, given that the aggregate supply curve is probably flat as a pancake right now the rise in inflation expectations suggests that NGDP expectations probably have risen sharply.
Has this been reflected in current performance?
The problem is that the economy is facing an enourmous headwind right now given the fact fiscal stimulus is being unwound at a very rapid rate. The negative effect of this was estimated as being so severe that many were expecting a double dip recesson about now. Krugman generously posted Goldman-Sach’s estimates of the effects of this withdrawl over a year ago:
And there are other factors in the first quarter that Ryan Avent has addressed, such as the weather and the Tsunami in Japan.
There’s no reason to expect investment to be growing until we get much closer to capacity (especially in residential construction), regardless of the real interest rate. And needless to say government consumption and investment expenditures are acting as a huge drag on the economy right now.
So given these facts where should we expect to finding growth right now? In personal consumption expenditures of course. PCE is up at a 3.4% annual rate over the last two quarters. Expenditures on goods are up at a 7.0% annual rate over the last two quarters. Expenditures on durable goods are up at a 15.8% annual rate over the last two quarters. (In fact expenditures on durable goods ran 4.6% above its prerecession peak in the otherwise lackluster first quarter).
Must have something to do with that $5 trillion increase in household stock wealth.
“If I were to go to your house and piss mostly in your toilet and half on your floor and argue that the piss in the toilet minus the piss on the floor has increased so you should thank me, you would be pissed.”
Remind me not to invite you over my house.
P.S. I left a comment over at DeLong’s site as well (on Japan’s moderately successful QE from 2001-2006).
I am not surprised that you do not understand what money is. Keynes did not understand what money is and you follow in his footsteps. According to him (and you) money is cash balances. Money is what does not buy bonds. Then what do you use to buy bonds. Cowries? I hope that some day some common sense descends on economists. But I can see that that day is probably half a century off.
Almost the entire community of macroeconomists did not anticipate the crash. And as this discussion shows they do not understand what to do in its wake as well. Any other discipline thinking of itself as a science would have realiased that its fundamentals were wrong. Instead, we have in economics tired old men trotting out tired old theories.
Anway, I anticipate that another crash is not long in the making. Maybe that will make at least a few economists realise that the fundamentals of economics are wrong.
I have absolutely no clue as to why you think that I think that money is cash balances. There is no basis for this delusion in my post or anything I have written.
On QE2 I absolutely assume that the change in the money supply doesn’t matter. I note that Bernanke et al agree with me. Otherwise they would have bought 3 month T-bills. I believe that the effects of QE2 are due to reduced supply of 7 year treasury notes to the public and not due to the increase in the supply of high powered money (balances at the Fed) or any effect on the supply of money. When I say money, I mean either m1 or m2.
I have absolutely no clue what balances of paper US dollars in the US are and I don’t care.
In any case, cash is only money if it is circulated. Vault cash in banks isn’t part of the money supply. I honestly have no clue how much cash is out there (now including vault cash and US dollars outside of the USA). I don’t care. It has been known for almost my entire life that cash is not a useful measure of the money supply.
When I visit the USA I don’t always bother to get paper dollars — sometimes I just use my debit card. I’ve noticed that paper money is not strictly needed in order to buy things.
No need to wait. Iceland is a prime example. Compare with Ireland. Enough said.