Monetary policy ALWAYS works for the “wrong” reason
First the Keynesians say QE “won’t work” because in their flawed interest rate-oriented models it doesn’t work at the zero bound. Then the Fed does QE1 and the dollar plunges 4% on the news. Mere rumors of QE2 are enough to push stocks and commodities up sharply, as well as TIPS spreads. So now they are forced to argue that it works, but it works for the wrong reason, so it still doesn’t work. Or something like that. I.e. it works because it’s a signal for more expansionary future monetary policies, not because of more money today. Fine, but isn’t that arguing about how many QE transmission mechanisms can dance on the head of a pin? Does it work or not, that’s the question. (If you haven’t read my preceding post, do so first.)
But the bigger flaw in the Keynesian argument is that they don’t seem to realize that monetary policy always works through signaling, not just at the zero bound. Here’s Nick Rowe:
[Update: Tom Hickey asks: “So monetary policy boils down to central bank communications leading to expectations?“
My response: That’s 99.9% of it, yes!
Here’s an example. On January 3rd 2001 the Fed cut it’s target rate from 6.5% to 6.0%, which was a bigger cut than expected. Stocks soared 5% on the news. (Even more, as the announcement was not 100% unexpected.) Does anyone seriously think that stock investors give a damn about the fed funds rate over the next 6 weeks? Of course not! Obviously the market was reacting to the signal the Fed was sending about future monetary policy. The Fed was showing a surprisingly strong determination to mitigate the post-tech bubble slump. We did have a recession, but it was surprisingly mild. Unemployment peaked at a mere 6.3% and we never saw even two consecutive quarters of falling RGDP.
If you don’t believe me, consider the following. While short rates fell slightly, long term bond yields rose dramatically on the news. The income and inflation effects at work. And long term interest rates are presumably of more interest to stock investors. Thus it wasn’t lower rates causing the stock surge, it was higher expected NGDP (and RGDP) growth, triggered by expectations of a more expansionary future monetary policy.
I suppose people like Woodford would argue that although long term rates (and hence future expected short term rates) rose, it was still expansionary because future rates fell relative to some sort of Wicksellian equilibrium rate (which was rising even faster.) Fair enough, but then if we only know this because we observe expected NGDP growth surging, then why not cut out the middleman? Why not stop talking about interest rates entirely, and use the expected growth rate in NGDP as our monetary indicator?
Regardless of how you feel about interest rates, it’s clear that monetary policy is always about signaling, not current instrument settings. So if the Fed is impotent at the zero bound because policy signals are unreliable, then it’s 99% impotent when not at the zero bound. There’s nothing special about the zero bound. The January 3rd 2001 rate cut would have done almost nothing to slow the rapidly oncoming post-tech bubble slump, were it not for the Fed’s ability to credibly signal future policy intentions.
Many people don’t know this, but the Fed also cut rates sharply after the October 1929 stock market crash, and yet (unlike 2001) NGDP and output plunged very sharply in 1930. There could be many reasons for this difference, but surely the fact that we were pegged to gold in 1929 made it harder for the Fed to credibly promise a policy of open market purchases a l’outrance to arrest the slump.
Here’s another example. Most economists think dollar devaluation in 1933 boosted NGDP. If you asked them why, they’d say something about improving the trade balance. In fact, the trade balance got “worse,” as the rapidly growing US economy sucked in more imports. Once again, the policy “worked,” but for the “wrong reasons.” The rate cut boosted expectations of the future money supply, and this boosted current AD for reasons explained in the standard Woodford/Eggertsson model.
The other complaint is that the signalling channel may fail to work for “credibility reasons.” That’s a completely phony argument, which I’ll address in my next post.
Tags:
6. September 2012 at 07:47
Scott,
But you always say expectations have to be about something. There has to a transmission mechanism. Just saying there is won’t convince Wordford, et al.
6. September 2012 at 07:56
Expectations make perfect sense when there are meaningful future policy actions to expect. When interest rates aren’t at the ZLB, there is still a lot of leeway for future action, so the signalling effect is very powerful. When rates are at the ZLB, according to Woodford the only thing you can get the market to expect is low rates for an extended period; this seems like an extremely weak policy lever, and is unclear if it’s politically possible to even commit to something like that regardless.
6. September 2012 at 08:06
4.5% NGDPLT
Signalling goes away entirely right?
Expectations are normalized, and that’s the REALLY radical thing, suddenly from here to eternity we know exactly what the NGDP # is in nay given quarter within +/- a couple tenths of a percent, with a confidence of 95% of better.
So once agan the question I can’t get an answer to:
ASSUME you get no make-up, but tomorrow we are at 4.5% forever.
Isn’t that change in assumptions FAR MORE powerful than adding a couple points in make-up?
6. September 2012 at 08:35
Current market conditions are a good example of expectations being 99.9% of monetary policy. Consider this. After the May payroll report came out in early June (it was bad), the market dropped substantially then rebounded a day or two later. It has stayed elevated since then. Why? Because everyone in the market knows that if inflation is low and employment is bad, then the Fed will step in with more QE and asset prices will rise. The market believes this because this is exactly what the Fed has been saying, over and over and over. The Fed has been crystal clear about this point and just like Pavlov’s dogs, the market has responded accordingly. Now, imagine what would happen if the Fed was crystal clear that it would not tolerate NGDP growth less than, say, %6? Without any doubt, we would see NGDP growth of 6% for as long as the Fed maintained that position.
6. September 2012 at 08:36
‘…but the Fed also cut rates sharply after the October 1929 stock market crash, and yet (unlike 2001) NGDP and output plunged very sharply in 1930.’
The discount rate, to be specific (I don’t know if there even was a FF rate back then).
6. September 2012 at 08:45
Scott, if the Fed came out tomorrow and said, the base is permanent and short rates will be zero UNTIL NGDP is back to its pre-crisis trend, what would you expect to happen to the spread between 30 year long bond yields and 10 year T note yields? I would assume they would go vertical.
What say you?
6. September 2012 at 09:10
There is an easy solution to this problem.
The issue is whether or not the Fed is depending on the banking system to lend money as a means to increase NGDP. At the zero bound, and given the assumption that the Fed depends on bank lending to increase NGDP, then the zero bound does present a barrier.
But, if the assumption of dependency on bank lending is dropped, then Keynesians would probably understand that the Fed can increase NGDP even at the zero bound.
Ask Woodford or Krugman if they will end up spending more money if the Fed purchased the trash they leave on their curbs each week. Ask them if they would hoard the cash they get, or spend the cash they get. That will probably give them the inkling they need to understand what market monetarists have in mind when they say the Fed can increase NGDP even at the lower bound.
6. September 2012 at 09:12
And of course, definitely, no question, the Fed would end up paying “the going market price” for the empty aluminum cans and beer/wine bottles. Even if the price rises to $5.00 a can and $8.00 a bottle, those are “going market prices.”
6. September 2012 at 09:18
Brito, So now the argument isn’t “signaling is weak,” it’s “signaling is weak at the zero bound.” Fine, Nick Rowe’s newest post addresses that issue. Stop targeting interest rates. Stop steering the economy with a steering wheel that locks up when you need it most.
6. September 2012 at 10:35
Scott says…
“The other complaint is that the signalling channel may fail to work for “credibility reasons.” That’s a completely phony argument, which I’ll address in my next post.”
I hope your argument does not have elements of the repubs suddenly declaring “opposite day” on monterey policy after the election.
6. September 2012 at 10:35
Professor Sumner,
What about Sept 2008, what was the Fed’s signal then? Beckworth and yourself have both argued that the Fed made a horrendous error not cutting rates that month, yet the market shrugged it off and was essentially unchanged that week and stayed flat until late Sept.
Keep in mind that week in Sept. was arguably the most tumultuous of the entire financial crisis — Lehman went into bankruptcy the weekend before, AIG was bailed out and deposit insurance was extended to the MMMF’s to halt mass withdrawals.
I’d say the markets were unconcerned with the lack of Fed action at the time, but still fairly confident the Fed could use the interest rate tool if necessary. The markets didn’t really panic at the lack of Fed action until AFTER the adoption of IOR (Louis Woodhill deserves credit for this argument). The markets, in my view, correctly foresaw that IOR would be a tool that greatly dampened any expansionary moves by the Fed — like applying constant pressure to the brake pad while trying to tap the accelerator.
With that in mind, shouldn’t IOR be removed as a precursor to further monetary expansion, under a ‘let’s-see-what-happens if..’ approach?
6. September 2012 at 10:39
“Brito, So now the argument isn’t “signaling is weak,””
That was never the argument as far as I’m concerned, at least among New Keynesians.
“it’s “signaling is weak at the zero bound. Fine, Nick Rowe’s newest post addresses that issue. Stop targeting interest rates. Stop steering the economy with a steering wheel that locks up when you need it most.”
I prefer we just cut out the middle man, change the system and end any debate about policy ineffectiveness as you know: http://neweconomicsynthesis.wordpress.com/2012/09/06/two-pieces-of-the-puzzle/
6. September 2012 at 10:55
I guess I should also mention the reason I believe understanding what happened from Sept. 2008 onward is important.
A commonly held view largely believes that the financial crisis created a sudden unusual increase in base money demand and that the Fed was caught off guard with this influx in demand. Tyler Cowen has made this type of argument — that in the ‘post-Lehman’ world, changes in money demand, banking and liquidity preference had become an unsortable mess.
The problem is that market data doesn’t bear this out. In Sept. 2008 with the financial crisis at its peak, the stock market was showing signs of recession, but not sharp economic collapse. The currency markets were not yet showing sharp appreciation of the dollar and major currencies. Commodities and TIPS prices had yet to plunge and nominal interest rates on debt securities were held in check.
All this points to the fact that the financial crisis was not the decisive factor that put the ‘Great” into the “Great Recession” — monetary policy was the culprit. Elevated money demand came afterward (especially post-IOR), in reaction to policy.
6. September 2012 at 11:18
“But the bigger flaw in the Keynesian argument is that they don’t seem to realize that monetary policy always works through signaling, not just at the zero bound.”
Nonsense. The NK model says the whole curve matters. Period. It feels like you’re so determined that the rest of the world is crazy that you are constitutionally incapable of actually acknowledging their arguments. This is ridiculous straw man.
6. September 2012 at 11:24
I need an answer here:
ASSUME you get no make-up, but tomorrow we are at 4.5% forever.
Isn’t that change in assuming constant level growth FAR MORE powerful than adding a couple points in make-up?
6. September 2012 at 11:39
Weren’t Keynesians some of the first to initially call for QE in the absence of satisfactory demand?
6. September 2012 at 12:42
Scott
Long term bond yields rose? I don’t see it in the data.
http://www.treasury.gov/resource-center/data-chart-center/interest-rates/Pages/TextView.aspx?data=reallongtermrateYear&year=2001
Also, will you admit that your monetary policy mechanism is through asset prices? None of that hot potato stuff?
6. September 2012 at 14:42
“The rate cut boosted expectations of the future money supply, and this boosted current AD for reasons explained in the standard Woodford/Eggertsson model.
The other complaint is that the signalling channel may fail to work for “credibility reasons.” ”
This reads to me like everyone agrees that monetary policy can boost activity at the ZLB by credibly commiting to a permanent lift in the money supply.
As a result, any difference in opinion is about:
1) How this credibility works (how do individuals, institutions, form expectations)
2) Whether commiting to this path is reasonable
Ignoring the second point (assuming that it is reasonable), the key difference is that you and Nick are stating that central banks can influence expectations of “medium-long) term future real rates of return – to me there is an implicit “multiple-eqm”/”co-ordination game” approach sitting in all of this that describes the difference between you view and the view put forward by standard interest rate targeting.
6. September 2012 at 16:34
I am an ardent Market Monetarist, and 100 percent support the idea the Fed should be transparent and explicit in setting NGDP targets and the methods they will employ to get to announced targets.
But I contend when the Fed buys Treasuries, the sellers have to do something with the money. They can spend the money (good), they can buy other assets (good) or they can put the money in the bank (not so good, but eventually the banks have to lend the money out, and IOR is important in this regard). So QE works, even with a obfuscatory, mysterious (and self-exalting) Fed, as we have now. If the Fed would just stick to QE in large enough proportion, it will work (either that, or human nature has changed, and nobody will spend money no matter how much of it is around).
I wonder what minute fraction of the population could even pick Bernanke out of a police line-up, or identify if the Fed is “loose” or “tight” right now.
QE will work, as Milton Friedman pointed out in his articles on Japan.
An interesting question is how to handle soaring Fed balance sheets, and the interest payments to the Treasury, Seems we have a chance to help deleverage th nation, and lessen tax burdens.
6. September 2012 at 20:41
Benjamin Cole.
What you say is true.
MF,
Also true. Bank lending doesn’t have to increase for people to spend the money the Fed prints to buy assets. The Fed can even buy standard vanilla T-bills which are 1 month and zero coupon by the way, for more than their market value. (Pay 1010 for a 1000 dollar bond)
7. September 2012 at 06:18
Scott,
One reason that most macroeconomists (particularly business/ Wall St. economists) are so wedded to framing monetary policy in terms of interest rates is that they focus so heavily on the expenditure components of GDP. If, for example, the Canadian economy is hit by a negative AD shock (say, weak US AD) and the BoC cuts rates in response, these economists will expect an increase in activity in domestic “interest rate sensitive” sectors of the economy (eg. housing) to offset the weak external demand. And if the BoC does its job correctly, that’s what these economists will observe: weakness in net exports and strength in housing. These economists then conclude that the primary channel by which monetary policy affects the economy is through lower interest rates. They also see a secondary channel coming through the impact of a weaker real exchange rate on net exports. “Expected future nominal income growth” is difficult to observe, difficult to trace through the expenditure components and therefore less appealing to most of these economists.
8. September 2012 at 07:19
[…] framing technique is that the same could be done with the 2001 interest rate cut I described in this post a couple days ago, which immediately boosted stock prices by 5% and helped prevent a deep post-tech bubble slump. […]
8. September 2012 at 10:06
Ritwik, Wrong series, you linked to real rates.
8. September 2012 at 10:23
Jor, Market monetarists were the first. Krugman was skeptical about the idea in early 2009. I asked him to call for QE, and he refused.
K, You said;
“Nonsense. The NK model says the whole curve matters. Period. It feels like you’re so determined that the rest of the world is crazy that you are constitutionally incapable of actually acknowledging their arguments. This is ridiculous straw man.”
That was poorly worded on my part. I meant that they make an argument against QE based on the zero bound, that would also apply when we aren’t at the zero bound. But at times they seem to imply otherwise.