The real problem with Fed policy
Five year TIPS spreads are at 1.5%. Because the Fed targets PCE inflation, and because 2.0% PCE inflation is roughly equivalent to 2.4% CPI inflation, the Fed’s target is roughly a TIPS spread of 2.4%. So 2014-19 inflation expectations are 0.9% below target. Here’s Ryan Avent:
American markets are once again hunkering down for a bout of disinflation. Expectations for inflation over the next five years have fallen half a percentage point since July, to around 1.5%: a level at which the Fed has previously moved to begin new asset purchases.
It’s important to recall that the Fed has a dual mandate, so this fact doesn’t necessarily imply that money is too tight. Later we’ll see that it is too tight, but let’s first consider the counterargument.
The Fed’s dual mandate covers both inflation and employment. Thus the Fed should push inflation above their target when unemployment is high, and they should push inflation below target when the unemployment rate is low. It seem likely that unemployment will be below average over the next five years, if only because it’s been above average for the previous six. So it’s possible that money is not too tight. Possible, but extremely unlikely. Here’s Ryan again:
THE monetary economics of a world in which interest rates are close to zero are not especially mysterious. Stimulating the economy at that point requires central banks to raise expected inflation. Disinflation, by contrast, results in passive tightening, since the central bank can’t lower its policy rate and since the real interest rate is the policy rate less expected inflation. In this world, the downside risks are much larger than those to the upside. There is infinite room to raise interest rates if inflation runs uncomfortably high (one might even welcome that opportunity to push rates up as that would reduce the probability that rates would fall to zero again in future). But there is no room to reduce interest rates if inflation is running to low. That, in turn, forces central banks to use unconventional policy or run psychological operations to try to boost expectations. Central banks are not very good at those sorts of things.
Suppose that the Fed runs inflation at 1.5% over the next 5 years, and then we hit another recession. In theory, the Fed should then push inflation much higher. But as Ryan’s comment suggests, exactly the opposite is likely to happen. The Fed will let inflation fall even below 1.5% in the next recession, and we’ll be in exactly the same position we are today.
This means that the real problem is not that money is too tight today (although it is) but rather that the entire monetary regime is flawed. Level targeting of prices would be better (and is the no-brainer solution to the euro-crisis, given their fixation with inflation targeting. But the ECB is not a no-brain, it’s a negative brain.) Another solution is NGDP targeting. Even better would be NGDP level targeting. These are ways of moving away from the Fed’s current procyclical monetary policy.
Ryan’s comment also points to the danger of passive tightening. Many people still have trouble with the notion that monetary policy could be tightening even as central banks “do nothing.” But clearly they can (and this isn’t just a MM view, it’s also a New Keynesian view, and an old monetarist view.)
TravisV sent me an article indicating that the Fed is beginning to understand the situation:
St. Louis Fed President James Bullard told Bloomberg TV that the Fed should consider delaying the end of quantitative easing in response to tumbling inflation expectations.
His concern was tumbling inflation expectations. . . .
Bullard was basically echoing the concerns of San Francisco Fed President John Williams, who suggested the Fed may have to increase its asset purchase program.
I don’t always agree with Bullard, but to his credit his views are always data driven. He’s an important swing vote at the Fed, as he’s one of the moderates.
PS. Once again, we’d have a much better idea of whether money is too tight if we had a NGDP futures market. But the Fed isn’t willing to spend $2 million dollars to set up a subsidized prediction market that would provide useful forecasts, even as trillions of dollars of wealth (and lots of potential jobs) are being wiped out each week. And the economics profession is equally apathetic. Over at Econlog I have a related post.
Update: Mark sent me an excellent and somewhat related post from a few days back by Tim Duy. He’d make these points even more forcefully today. And take a look at the dovish shift in the FOMC next year. I’d wouldn’t be at all surprised if there were no fed fund rate increases in 2015. Where are 4 votes for higher rates? The real problem is the eurozone.
Tags:
16. October 2014 at 08:26
> “So 2014-19 inflation expectations are 0.9% below target.”
Inflation expectations are 0.9 *percentage points* below target, which means inflation expectations are 38% below target! Let’s use the bigger numbers to be meaner to the Fed, it could make it easier for MM-sympathetic journalists to write catchy headlines.
Also, this post obviously illustrates the ways in which “money is tight” can have multiple correct meanings. On the one hand there’s your usual “money is tight = NGDP below target.” On the other hand you’re willing to use “money is tight = Fed policy is undershooting its existing dual mandate” in this blog post.
To poke you a little bit, I might argue therefore that the phrase “low interest rates = money is loose” is not incorrect, just bad in a normative sense. As with “monetary offset,” you like to exploit a blurred line between normative and positive statements, and I’m going to be a jerk and insist that we distinguish between them. 🙂
16. October 2014 at 09:17
Good stuff. I think money is too tight. I hope Yellen reads your blog.
16. October 2014 at 09:26
Great (related) post by Tim Duy:
http://economistsview.typepad.com/timduy/2014/10/the-methodical-fed.html
16. October 2014 at 09:35
Mark, excellent link.
16. October 2014 at 09:52
To what extent do declining market based measures of inflation simply represent a mechanical response to lower oil prices? The drop in oil is not just demand driven but supply driven. Libyan oil production has tripled over the last several months. Is US corporate pricing power really receding as quickly as financial markets are pricing? Given the latest earnings data, I find that position highly dubious.
16. October 2014 at 10:01
TD, Good point, but some vagueness is inevitable.
Mark, Thanks, I added an update.
Neil, It’s certainly partly oil (a point I made in an earlier comment section), but the sharp fall in stocks and long term nominal bonds yields is essentially 100% unrelated to oil. So there’s more than oil involved here. And Libya’s too small to explain the big oil price drop.
16. October 2014 at 11:27
I do wonder how much the Fed is thinking about market-based measures of inflation. After all, five year forward breakevens were even lower in 2011 and in the September 2011 FOMC policy directive, longer term inflation expectations were declared to be “stable”. I doubt monetary policy officials will flip as quickly as markets, which is why ending the QE program should remain the basecase for October. We’ll see!
16. October 2014 at 11:41
What is the impact of the Euromess in the US? Does European collapse passively tighten policy in the US?
Is the fall in oil prices driven by reduced worldwide demand for oil?
16. October 2014 at 12:34
Maybe all this market turmoil is good news. It will hold the fed from tightening too early and heeding the insane calls of the hawks
16. October 2014 at 12:49
Scott, you may have already discussed this in the past, so my apologies if I’m rehashing old ground. But just saying monetary policy is too tight right now needs (for me at least) an explanation of *why* it is too tight given a zero interest rate and a large federal reserve balance sheet six years after the financial crisis. Are you assuming a pseudo Keynesian model where the propensity to save is higher than the amount of profitable investment even when the federal funds rate is zero? It sounds like you and PK have the same underlying model of what’s wrong with the economy (too much saving and not enough investment), but your solutions are different (monetary vs. fiscal policy). You’ve said that fiscal stimulus would lead to misallocated capital (at least I think you have). Wouldn’t negative real borrowing rates lead to misallocated capital also (assuming the Fed could drive inflation expectations significantly higher)? Maybe we should be looking at the source of the underlying problems instead of just trying to address the symptoms with monetary and/or fiscal policy after six years.
16. October 2014 at 12:58
Yes I would make the point that some of the fall in inflation isn’t bad for NGDP. Falling oil and strong dollar are partially supply side phenomena and indicating stronger growth.
Only thing flashes red is the 10% Stock correction but those happen from time to time. And may partially be a result of the stock market not fully incorporating NGDP targeting as past problems occurred when TIPs were signaling a problem, but at those times it was NGDP falling this time evidence points to some of the TIPs move as being supply side oil related.
But the FED targets inflation not NGDP so for that reason they should be moving more dovish even though oil prices shoudln’t be what they target.
Sometimes I think the whole crisis was caused by the FED misinterpreting $150 oil instead of missing the housing crisis.
16. October 2014 at 13:13
The real problem of Fed policy is that it refuses to accept the truth that the disinflationary/deflationary forces that seem to be significant despite the last few years of significant inflation, is that the market is sickened by too much inflation! The market is telling us, and the market is always the most right, that what it needs is significant deflation to purge the economy of malinvestment.
Does that mean I am asserting that if you ask anyone on the street “Do you welcome widespread unemployment?” that I am claiming that yes, yes they do want that, then you have not grasped the point about what the market wants. Does one have to claim that the man on the street wanted the candlestick makers to lose their jobs in the face of competition from electricity? Or that the typewriter workers should lose their jobs from the onset of personal computers and home printer production? Of course not! So don’t tell me that a desire for market driven deflation to run its course, despite the temporary unemployment that arises, to be some ham fisted, disconnected, disinterest in standards of living.
The world is such that when people are misled, oftentimes the deceivers are not in a position to fix the mess they created. That the specific type of deatruction they unleashed can only be fixed by the victims left on their own. Stop believing in Santa Clause. Charge the guilty Federal Reserve agents with crimes against humanity, and give the market room to breathe life back into an almost completely warped economic system.
Have any of you true believers ever stopped and asked yourself what would the economy have to have looked like if all that money printing you desire and say is not enough, was positively destructive and not “it is insufficient” destructive? Of course you haven’t! To even ask that question would compel you as “intellectuals” to dig and dig until at last oops, you might end up discovering that the concept of malinvestment applies to your intellectual pursuits. The risk of cold hard reality crushing your monetary socialist dreams would be too psychologically disruptive to handle. Better to stay asleep and dream that the philosopher kings will find a way to use the very poison that hurts us, at a different dosage rate, much like bloodletting at a better rate has to be the solution to the problems caused by bloodletting.
A few years ago on this blog I predicted that in order to prolong the malinvestment and prevent market driven deflationary correction, that the money supply would have to accelerate. We’re still confirming that to this very day. Now the mere rumor of the end of QE, which is itself an acceleration, has resulted in the market going into the correction it has so desperately yearned for since 2008, and much earlier. Stats are plunging all over the world. The market has Goodharted the new, higher rate of money printing.
You’re all digging your own absolutist authority caused graves, exactly like the socialists (communists) of the 19th and early 20th century. You don’t want to accept the reality of mankind. Gods destroy until they are themselves destroyed.
16. October 2014 at 13:19
Michael Byrnes — I think we would say ECB policy is a negative shock to the US and world economy (and a totally needless one, of course).
Of course the big story in oil is US production, but I’d keep an eye on Chinese demand.
16. October 2014 at 13:26
TD — the problem is that “tight” and “loose” are always relative to something, but that something is usually unstated.
I think generally if we say “money is too tight” we mean “a less tight monetary policy would have led to better employment, better real GDP growth, etc.” that is to say the word “too” implies we are comparing actual policy to ideal policy. But even then, the “etc” can have a lot of ambiguity, especially as regards things like the desirability of price stability. And then there’s the question of what inflation even means… okay, that’s enough monetary policy for me today.
16. October 2014 at 13:46
Scott,
I understand all the arguments about the Fed missing their inflation target low. But do you really truly believe that 1.5% inflation means a lot more jobs than 2% inflation? If so, how long will the stimulative effect of 2% inflation last on the job market? If the job market doesn’t do well at 2%, should they increase their target to 2.5% or 3%? I just can’t see the direct causal link between hitting 2% instead of 1.5% and a stronger economy. By what mechanism is this supposed to work?
16. October 2014 at 14:02
@John
Go back and read Scott’s archives. It’s not about inflation, it’s about NGDP growth. Start here http://www.themoneyillusion.com/?p=20433
16. October 2014 at 16:23
Excellent blogging.
You know, the penalty for a monetary policy that is a little too loose is mild inflation at full output.
THe penalty for a monetary policy that is too tight is a ZLB permanent recession (see Europe and Japan).
is this a difficult choice to make?
Do people reaaly prefer recessions to mild inflation?
16. October 2014 at 18:10
Why are most economists really concerned with monetary policy if they think money is neutral? Are they highly concerned over the short term?
16. October 2014 at 19:56
CMA, only Ed Prescott would argue that money is neutral in the short-run (and the people that know him well often hint that he doesn’t really believe this either).
If money is non-neutral in the short-run, then monetary policy matters. And everyone agrees that money is non-neutral in the short-run.
16. October 2014 at 20:18
I know money is supposed to be neutral long term and short term non neutral. That’s why I ask if everyone is worried over the short term. In the long term it doesn’t matter anyway if neutrality is true.
16. October 2014 at 23:22
benjamin cole:
The choice you’ve set up cannot be answered rationally until you have substantiated the theory that insufficient inflation is responsible for “permanent recession”. You have not done so. Why hasn’t the market adjusted prices, especially relative prices, in Japan after all this time? If you believe as Sumner does that ANY inflation rate will eventually be adjusted to, then your assessment of Japan is, to that extent, untenable.
16. October 2014 at 23:31
TallDave:
“I think generally if we say “money is too tight” we mean “a less tight monetary policy would have led to better employment, better real GDP growth, etc.”
Imagine a Keynesian telling you that when he says “deficits are too small”, he means “a greater government budget deficit would have led to better employment, better real GDP growth, etc.”
I am sure you will agree that this form of argument is flawed
17. October 2014 at 04:59
> Level targeting of prices would be better (and is the no-brainer solution to the euro-crisis, given their fixation with inflation targeting. But the ECB is not a no-brain, it’s a negative brain.)
I used to think so too, but looking at the actual price level data (http://research.stlouisfed.org/fred2/series/CP0000EZ18M086NEST), I don’t think it’s such a no-brainer anymore. From inception (Jan-1996) up until Sept-2007, the price level target was kept spot on (25.9% HICP increase in 140 months, or 1.99% inflation per year). After Sept-2007 they seriously lost control. However, almost half the time there was overshooting (Sept-2007 – Nov-2008, Oct-2010 – Feb-2013), so while they’ve been undershooting their target since early 2013, the cumulative price gap starting from Sept-2007 is not all that large: HICP level was 117.7 August 2014, while hitting a perfect 2% target starting from Sept-2007 would yield 120.1%, a cumulative undershooting of 2% over 7 years.
So while I think you can make a case a good case for a price level target, especially when placing more weight on the recent past and on expectations, it does not seem like a complete no-brainer to me. During most of the 2011-2013 period the realized price level was actually above a level target.
17. October 2014 at 04:59
> Level targeting of prices would be better (and is the no-brainer solution to the euro-crisis, given their fixation with inflation targeting. But the ECB is not a no-brain, it’s a negative brain.)
I used to think so too, but looking at the actual price level data (http://research.stlouisfed.org/fred2/series/CP0000EZ18M086NEST), I don’t think it’s such a no-brainer anymore. From inception (Jan-1996) up until Sept-2007, the price level target was kept spot on (25.9% HICP increase in 140 months, or 1.99% inflation per year). After Sept-2007 they seriously lost control. However, almost half the time there was overshooting (Sept-2007 – Nov-2008, Oct-2010 – Feb-2013), so while they’ve been undershooting their target since early 2013, the cumulative price gap starting from Sept-2007 is not all that large: HICP level was 117.7 August 2014, while hitting a perfect 2% target starting from Sept-2007 would yield 120.1%, a cumulative undershooting of 2% over 7 years.
So while I think you can make a good case for a price level target, especially when placing more weight on the recent past and on expectations, it does not seem like a complete no-brainer to me. During most of the 2011-2013 period the realized price level was actually above a level target.
17. October 2014 at 05:40
Neil, We know they look at them.
Michael,
1. Significant
2. Yes.
3. Yes.
Edward, Unfortunately, the markets are predicting that the market turmoil will not be good news.
Mark, Interest rates and the Fed balance sheet are not good indicators of whether money is tight. That’s not just my view, it’s the standard textbook view. I’d encourage you to look at some of the papers that explain my views in more detail, links in the right column of this blog.
I am not advocating lower real interest rates, just faster expected NGDP growth. So I do not have the same model as Krugman.
Sean, Maybe, but the safer assumption is always that stocks are rational. I don’t believe oil is falling for supply reasons, at least not in the past couple weeks. On the other hand, in my view the fall in US NGDP expectations (so far) has been quite modest.
John Becker, There are several misconceptions in your question. I compared 1.5% inflation to 2.4% inflation. I said the Fed has a dual mandate. There’s a huge difference between increasing inflation to try to hit a pre-announced target (generally a good idea) and raising the target (generally a bad idea.) Whether higher inflation affects jobs depends where it comes from. It does create lots of jobs if it comes from the Fed during a period of slack. It doesn’t if it comes from the supply side, or during a period of no slack.
Ben, You said:
“Do people really prefer recessions to mild inflation?”
Ask that question in German.
CMA, The short run effects of tight money caused the 1929-33 slump, which caused 25% unemployment and put the Nazi’s in power. Does that concern you?
MC, People have drowned in lakes that averaged 3 feet in depth. It makes no sense to have inflation run well above target when unemployment in low 2007-08, and then run well below target when unemployment is high. It should be the opposite. And they are idiotic to use HICP as their price index. That’s another choice the ECB got wrong.
Under a price level target inflation would have been less volatile over the past 10 years, as markets would have known the ECB would have to bring prices back to the trend line.
17. October 2014 at 06:05
Mark,
There is one thing professor sumner did not clean up in his response to your comment. I’m going to do my best to put words in his mouth, I hope he’ll correct me if I’m wrong.
You paraphrased professor sumner as saying fiscal stimulus was a bad idea bc it leads to ‘misallocated’ capital. While I believed he DOES think fiscal stimulus, especially here in years US, leads to wasteful spending, this IS NOT his important arguement against fiscal stimulus. Professor sumner has claimed, pretty successfully in my opinion, that fiscal stimulus is useless because we’re are in fact mostly getting the ngdp level path that the fed wants. This is the wrong path. But if fiscal stimulus succeeded in raising it, professor sumner contends that the fed would ‘tighten’ (or be less loose) in response and undo the good. The waste part is separate from this.
17. October 2014 at 07:32
Major freedom continues his insanity. He would have a relevant point about candle makers and horse drawn carriages being unemployed were he to show the new jobs being created in a demand side recession to replace the old ones. Positive supply shocks feel very different than negative demand shocks. That is the point that gets lost in his lunatic ravings.
17. October 2014 at 07:34
Next up: “you have not showed what I said to be ‘lunacy’ ”
So predictable!
17. October 2014 at 10:13
Scott, thank you for addressing my comment. And thank you Nick for clarifying a key point as well. I have in the past read some of the materials on the right side of your blog. Admittedly, I didn’t understand your views as well as I do now thanks to your and Nick’s responses.
If I understand your view correctly, it goes something like this (this may be wrong, so please correct me if it is): most business cycles result from demand side fluctuations in nominal spending spending which bump up against frictions such as sticky wages and prices. Monetary policy can smooth out the effects of these imbalances by stabilizing nominal spending (assuming it is always possible and credible). This would minimize fluctuations around trend growth. If NGDP (or NGDP level) targeting is achievable, and if it can be shown that it increases welfare over holding prices stable in the long run, then I would agree. Though I am not yet convinced it is achievable and would necessarily increase welfare in the long run.
Even if I’m wrong it still doesn’t make sense to me that there would be such a divergence between what was tight vs. loose monetary policy compared to prior to the financial crisis. You can’t ignore the microeconomic details of large firms sitting on stockpiles of cash and not investing it, and consumers wanting to save more than they currently are able to.
The only remedy I see for this is stronger productivity growth. But the economy can’t (sustainably) grow faster than the underlying fundamentals dictated by total factor productivity, and the quantity and quality of capital and labor. I don’t see how NGDP targeting would lead to stronger growth in these fundamentals. And if doesn’t, I don’t see how NGDP targeting would necessary be achievable.
Thanks again.
17. October 2014 at 22:13
Do people really prefer recessions to mild inflation?
One might ask: “which people?” The Great Depression wasn’t that bad if you had a job, or had invested very conservatively.
But the economy can’t (sustainably) grow faster than the underlying fundamentals dictated by total factor productivity, and the quantity and quality of capital and labor. I don’t see how NGDP targeting would lead to stronger growth in these fundamentals.
I think the simplest way to understand this is to imagine what would happen to investment if the Fed announced they were going to target negative 5% NGDP growth. (Better yet, assume you have advance knowledge that this is going to happen next Tuesday — how do you invest?)
And then consider that is essentially what the Fed did in 2008 and the consequences, and the picture starts to become clearer. Add in the period over which the ECB and BOJ have essentially targeted zero NGDP growth, and the case becomes empirically compelling.
18. October 2014 at 06:44
Mark, you said,
” I don’t see how NGDP targeting would lead to stronger growth in these fundamentals. And if doesn’t, I don’t see how NGDP targeting would necessary be achievable.”
What do real variables and TFP have to do with nominal aggregates? Presumably if you feel that NGDP targeting would not be achievable because of real factors, you feel the same way about an inflation target? This is completely at odds with the literature on monetary policy, so what do you base this on?
18. October 2014 at 09:13
Hi Ben,
Your point is valid. But what I’m suggesting is that it is much harder to, say, push inflation expectations to 4% (assuming a NGDP target of 5% and an expected real growth rate of 1% for some time) than to keep inflation from dropping too far below 2%.
In terms of real variables and nominal aggregates, you can’t ignore the microeconomics of individuals and businesses choosing prices and wages in a particular economic environment. I realize many people on this blog don’t like to think about NGDP as real GDP plus inflation, but that’s what it is (even if inflation is determined ex post–and even if our measures of inflation aren’t perfect). To promise to push inflation significantly higher in the future means real wages have to rise and for that to happen unemployment has to come down and labor force participation has to increase. Would higher inflation expectations necessarily accomplish this? Maybe, but I have a lot of trouble thinking through how the Fed could credibly promise to raise inflation in the future and keep it from falling if there are real deflationary pressures. If markets suspect they wouldn’t be able to maintain the level, then the expectations channel breaks down and we wouldn’t have higher NGDP now.
If I understand the theory behind NGDP targeting (and I may not) it works by reducing the imbalances in the labor and goods markets caused by demand side fluctuations in nominal spending bumping up against nominal frictions such as sticky prices and wages. But a (seemingly) perpetual negative wicksellian rate is more of a structural problem that seems to me wouldn’t be solved by NGDP targeting and, in fact, could make NGDP targeting not credible.
In general, my qualm isn’t about NGDP targeting though. Something like it–or price level targeting–might be a good idea. My qualm is about people thinking the economic conditions we’re facing now (and for the past six years) have all been because the Fed’s monetary policy is too tight.
Thanks for your reply.
18. October 2014 at 09:38
Just to follow up on my last comment:
The effects I see from NGDP targeting based on the theory as I understand it (which may be wrong) is that it reduces the fluctuations around a trend growth rate of, say, 2% per capita with a mean reversion (i.e. catch up growth after recessions). If there are structural problems or real shocks (seen or unseen), then I don’t see how a “looser monetary stance” would necessarily help.
Also, the empirical evidence I’ve seen suggest that inflation expectations are more adaptive than forward looking, which makes inflation targeting much easier than NGDP targeting. Thus, “…it is much harder to, say, push inflation expectations to 4% (assuming a NGDP target of 5% and an expected real growth rate of 1% for some time) than to keep inflation from dropping too far below 2%.”
Thanks.
18. October 2014 at 10:12
Mark wrote:
“If there are structural problems or real shocks (seen or unseen), then I don’t see how a “looser monetary stance” would necessarily help.”
It would help because structural problems and real shocks can have monetary consequences (if not addressed properly by the issuer(s) of currency) that can themselves have real negative effects.
Obviously an NGDP level target will not, say, produce oil during a massive oil shock. A massive (negative) oil shock is (as Scott Sumner has put it) a decline in real income. The fracking boom is the opposite – on net we are wealthier because of it. There is no getting around those real effects by adjusting the money supply. But those real effects are not the cause of recessions.
Scott has addressed the argument about whether real shocks cause the business cycle many times, my favorite of which is here:
http://www.themoneyillusion.com/?p=12362
18. October 2014 at 10:56
Good point Michael. Real shocks can lead to monetary shocks. But real (negative) shocks also create downward inflation pressure. So lowering the interest rate below the Wicksellian rate, while maintaining a 2% inflation target, should address this–though admittedly perhaps not as powerfully than lowering the interest rate and raising expected inflation. The NGDP target argument (as I understand it) is that it is less efficient to target inflation than to (assuming you can) target nominal spending. Not that inflation targeting doesn’t also smooth out monetary disturbances.
Really all I’m saying is that our wealth can only (sustainably) grow if our productivity grows. Real growth can’t be pushed by monetary policy no matter how good it is at smoothing out fluctuations. So when there appears to be issues with our productivity growth, the answer isn’t necessarily that our monetary policy is too tight.
Thanks.
18. October 2014 at 12:42
Mark,
Two words – STICKY WAGES.
Look it up.
18. October 2014 at 12:49
Mark wrote:
“So when there appears to be issues with our productivity growth, the answer isn’t necessarily that our monetary policy is too tight.”
No, but slow productivity growth should not lead to recession – we only think it does because it is often accompanied by poor monetary policy. Poor productivity means that the goods and services that people would like to buy are in shorter supply than they would otherwise be. If demand exceeds supply, the result should be inflation (the good, appropriate type of inflation), not recession.
“The NGDP target argument (as I understand it) is that it is less efficient to target inflation than to (assuming you can) target nominal spending. Not that inflation targeting doesn’t also smooth out monetary disturbances.”
If you believe inflation targeting is possible, there’s no reason to think NGDP targeting is not. The Fed could institute an NGDP level target tomorrow, relatively easily, by drawing a line on a graph (of NGDP vs. time), and annoucing that the first rate hike will happen when NGDP exceeds the target.
Inflation (and thus RGDP) cannot be measured more reliably than NGDP . Estimates of inflation and RGDP include whatever error exists in the NGDP measurement *and* whatever error exists in the inflation measure.
Beyond that, there are situations where prices *should* rise, such as severe commodity shocks. Under an inflation target, a severe oil shock that drives prices up would require the central bank to tighten – to force other prices downward to offset the rise in oil prices. So the response to a supply shock under inflation targeting would be to initiate a monetary shock. (This is part of why central banks use “flexible” inflation targeting – leaving themselves wiggle room to avoid doing this). Under NGDP level target, they don’t need to concern themselves with prices and inflation during supply shocks.
18. October 2014 at 13:26
Mark, Policy only looks “loose.” It is not. Rates are low because the economy is weak, not easy money. I predicted that the ECB’s attempt to raise rates in 2011 would fail, and make rates even lower, for even longer. I was proved right. (And yes, it’s counterintuitive.)
Sorry I don’t have more time now.
18. October 2014 at 13:32
Trying to explain market monetarism to its skeptics is kinda like trying to explain aerodynamic lift to someone who insists that friction doesn’t exist. Or, if it does, it’s irrelevant in practice – and the only reason it exists at all is because of the evil gub’mint.
Because that’s what you age when you apply Stone Age moral intuitions in the modern world.
18. October 2014 at 14:31
*that’s what you get
18. October 2014 at 18:43
Scott, Michael,
Thanks for your replies. I appreciate your thoughtful responses.
Scott, You were right about the ECB. I think most people agree that the ECB made a mistake when it raised rates in 2011. But that’s not just a MM view.
I’ve made a sincere effort to understand MM. I may not have the best grasp on it all, but I have (and still am) trying. So far I haven’t been convinced that it is: 1) feasible under all circumstances; and 2) optimal under all circumstances.
That’s not intended as a disrespect for your views. I’m here because I’m trying to learn more about MM.
When we’re talking about zero nominal interest rates, raising nominal spending (to me) means raising inflation higher and pushing the real rate lower. I have the same issue with this that I have with the electronic money theory (where rates can be negative). Lowering real rates well below zero is not conducive to productive investment. And *unproductive* investment is wealth destroying in the long run. So unless pushing the real rate lower now leads to a higher real rate (which it might), I don’t think it is helpful.
Technically, lowering the real rate further now means we would be trying to pull future consumption forward to the present. I think it could be problematic to spur (potentially unproductive) investment through a strongly negative real rate while trying to pull forward future consumption that may not be there to draw on. Perhaps that is a limited view. Maybe the Fed could raise NGDP through QE and a sustainable positive rate would eventually emerge. But I don’t know that it would necessarily work, or that it wouldn’t lead damaging effects on emerging markets and asset prices.
On a much broader level (and I hesitate to say this because I’m not necessarily wedded to this idea, and it may offend some people), periods of temporary elevated unemployment resulting from demand shocks weed out unproductive firms. In the long-run, it is helpful for productivity growth to have such adjustments. A sort of survival of the fittest mechanism if you will. That said, long periods of high unemployment (like now) are not necessarily helpful.
Michael, I’m not sure about your assertion that a productivity slowdown means people are not able to buy the goods they want and thus “good” inflation will occur. I think we’re talking we’re talking about different things.
Daniel, I do understand friction and at one point studied aerodynamics (though I couldn’t explain it to you now). I also understand sticky wages and prices. One might even call adaptive expectations “sticky inflation.” One might also believe the Philips curve is flatter on the low end of inflation. There are a number of ways one can speculate that it would be much harder for the Fed to stabilize nominal spending growth when real productivity growth fluctuates. And that it may be *very* hard when we’re stuck on the zero lower bound.
Perhaps one of you can help me with a question I have. Can someone explain (or point me to an explanation) of how a national NGDP futures market would work? I’ve read a little bit about the one Scott is working on now, but how would it work for, say, a U.S. NGDP market that the Fed uses to target NGDP levels? Specifically: What would be the payoff structure and how would it affect people’s spending?
Thanks to all of you for helping me to learn more about MM. Perhaps if I learn enough I’ll change my views. Sorry for such a long post.
18. October 2014 at 20:50
Mark, a couple of points.
You said I made a good point, but you didn’t address it. Again, when you say
“Your point is valid. But what I’m suggesting is that it is much harder to, say, push inflation expectations to 4% (assuming a NGDP target of 5% and an expected real growth rate of 1% for some time) than to keep inflation from dropping too far below 2%.”
Why would this be the case? There is no basis for this kind of difficult asymmetry in policy in the monetary policy literature. If it were true that nominal targets were somehow harder to hit in one direction, it would be obvious empirically, and would be one of the classic ‘stylised facts’ of macro. So what are you basing this on? Just your intuition?
Then there are a few more things you say that are confusing.
“To promise to push inflation significantly higher in the future means real wages have to rise”
Higher inflation means lower real wages.
“But a (seemingly) perpetual negative wicksellian rate is more of a structural problem”
In the models (Gali, etc) the Wicksellian rate is dynamic and moves over time, and it represents the effect of demand shocks, not structural shocks.
Why does “which makes inflation targeting much easier than NGDP targeting,” follow from “empirical evidence I’ve seen suggest that inflation expectations are more adaptive than forward looking,”? And what is this empirical evidence you’re reference?
“Real growth can’t be pushed by monetary policy no matter how good it is at smoothing out fluctuations. So when there appears to be issues with our productivity growth, the answer isn’t necessarily that our monetary policy is too tight.”
Everyone here agrees with you, but – as you say in your last sentence – its orthogonal to the question of whether money is too tight. The question of “Is monetary policy too tight” can be answered with no mention of productivity growth. Talking about productivity growth in this context is a giant red herring.
“Lowering real rates well below zero is not conducive to productive investment.”
Why? Is this just your intuition again? And do you think the ideal policy world of MM has negative real rates? In that case, you have drastically misunderstood. I don’t want to speak for Scott, but if you asked him “Why do you advocate for negative real rates” I bet 10 bucks he would answer something like “Actually, my preferred policy regime lowers the probability of incidences of negative real rates”.
“periods of temporary elevated unemployment resulting from demand shocks weed out unproductive firms.”
This is a proto-Austrian view that question-begs by defining recessions as episodes of creative destruction . But this is not true for nominal shocks. It is also empirically a very weak claim. The rate of formation of startups and hirings by startups is independent of the business cycle, implying that creative destruction is independent of the business cycle. I can go on, but the point is that your premise is false, and so your conclusion that recessions are purifying is also probably false.
I’m not trying to have a go at you, but I think you’re basing your understanding of monetary policy and the goals of MM on your intuitions about macro, which is distorting your understanding.
18. October 2014 at 20:58
Excuse my typos.
19. October 2014 at 07:01
Mark, I’m not just saying the ECB made a mistake in 2011, I’m saying it caused both the 2008-09 and 2011-12 slumps.
Central banks should never, ever, pay any attention to interest rates, nominal or real. Markets should set interest rates.
It is not true that easy money leads to lower interest rates, real or nominal, except perhaps in the very short run. But not over periods of time that matter. Rates in the eurozone are really low because money is really tight.
It’s certainly not true that easy money shifts spending from investment to consumption, indeed just the opposite. Monetary stimulus tends to increase investment more than consumption. Don’t even think about interest rates when thinking about investment (never reason from a price change.) Interest rates are not monetary policy, NGDP growth is.
I have a paper on NGDP futures targeting linked to in the right column.
You said:
“On a much broader level (and I hesitate to say this because I’m not necessarily wedded to this idea, and it may offend some people), periods of temporary elevated unemployment resulting from demand shocks weed out unproductive firms. In the long-run, it is helpful for productivity growth to have such adjustments. A sort of survival of the fittest mechanism if you will. That said, long periods of high unemployment (like now) are not necessarily helpful.”
Studies have shown just the opposite, creative destruction works best at full employment. The government should never, ever intentionally create lots of human suffering on the totally unproven and theoretically unsupported theory that destroying jobs and wealth now might make us stronger in the long run. The economy is not a guinea pig.
19. October 2014 at 10:05
Ben J., thank you for the long thoughtful response. You are correct, I didn’t explain my points very well. I’ll try here:
The reason I see it being harder to raise inflation now to 4% is because we can’t lower nominal interest rates. Under normal circumstances you lower the federal funds rate to spur inflation. I do understand the expectations channel that views it more from an expectation of future inflation causing more spending now. But I don’t think it is as straight forward as thinking we can set a nominal target and markets will magically take care of the rest.
I think of a Phillips curve where inflation, pi, is indexed by the previous period’s inflation (inertial). So inflation is equal to the past period’s inflation, pi(-1), plus some slope term, alpha, times the output gap, (y-y^), [where y^ is potential output], plus a shock term, epsilon.
So it looks like this:
pi = pi(-) + alpha*(y-y^) + epsilon.
The pi(-) term is an inertial force, the alpha*(y-y^) is currently a drag, and the Fed has been trying to overcome this drag by affecting the epsilon (shock) term through forward guidance and QE. So with a drag from the output gap and inertia from the pi(-) term, it’s easy to see how it could be hard for the Fed to repeatedly “shock” inflation higher and higher at the zero lower bound.
That said, if you replace the inertial term with a forward looking term, then you are correct. But what I’ve read suggests the empirical evidence is that inflation is adaptive (however, I’m sure there is continuing debate about this). You can google “adaptive expectations” or “inertial inflation” to see evidence of this.
Here’s one way to view current economic environment:
More or less rational households planned their lifetime saving and consumption based on there being roughly a 2% per capita real income growth, and that they could count on at least a 6% real return on their savings (in the stock market and/or in real estate). They assumed these figures were relatively fixed because it’s been the case for a long time.
To everyone’s surprise these assumptions turned out to be false (or it is highly suspected they may be false). It turns out that these numbers may have been too high, by as much as a factor of 1.5. To smooth lifetime consumption households have adjusted down their current *and* expected future consumption.
From a rational expectations perspective, lowering the real rate of interest pulls future consumption forward. However, under the story I outlined above, there is less future consumption to draw on, and their are limits of how much consumption you can pull forward due to the requirements of old age. And now, even at a nominal federal funds rate of basically zero, there isn’t enough consumption, present or future, to support sufficient investment unless productivity increases substantially and changes our rational agent’s plans back to what they were.
In this model, businesses aren’t refraining from large new investment opportunities because borrowing rates are too high, but rather because their isn’t enough future demand expected to incentivize these new investment.
For there to be adequate demand, households need to spend a larger share of their income, not just spend more in nominal terms. For households to significantly increase their consumption relative to their income, their *real* earnings have to rise. And the dollar amount of required productive investment going forward needs to exceed the savings rate (to support inflation). And the incentive to invest will be dependent on how much *subsequent* future productive investment opportunity there is because aggregate spending created by aggregate investment in one period will fall short of costs (because of saving) unless there is more *productive* investment going forward.
[I emphasize *productive* because investment in and of itself isn’t enough. If it’s not productive, it will ultimately destroy wealth rather than create it. I think we all probably agree on that.]
So from this perspective, lowering the real rate further is unlikely to spur a higher rate of consumption *relative to income*. Only an increase in real earnings will accomplish this which has to happen from real productivity growth.
That model doesn’t take into account some other forces that may be depressing demand as well. The current account is about 2% of GDP now, down from about 6% of GDP earlier in the 2000’s. If the U.S. begins to grow relative to the other industrial countries, then this percentage will likely go up. I’m not just saying the current account will go up in dollar terms, I’m saying it will likely go up in percentage of GDP terms. This would likely decrease demand.
On the income inequality side (which is controversial, I know), a higher percentage of national income going to people that save a larger proportion of their income (wealthy people) also depresses demand.
I (or the hypothetical households in my model) may end up being pleasantly surprised to find that we start to return to trend growth, and then we can go back to our old model of the economy with positive interest rates and the usual business cycle fluctuations around trend. If this doesn’t happen, then it may take many years more for things to change much.
I hope this very long response addressed most of your questions to me about real versus nominal, and productivity, etc. I started to address them one by one, but I thought it would be too long. I probably went on even longer, though, than that would have been in this response. Perhaps it is *unproductive* blog responding on my part. I hope Scott doesn’t get irritated with me.
Thanks again for your responses.
19. October 2014 at 12:18
raising nominal spending (to me) means raising inflation higher and pushing the real rate lower.
That’s just bullsh*t. Where’d you get than nonsense from ?
periods of temporary elevated unemployment resulting from demand shocks weed out unproductive firms.
Likewise, we should stop vaccinating children, to ensure only the fittest survive.
There are a number of ways one can speculate that it would be much harder for the Fed to stabilize nominal spending growth when real productivity growth fluctuates.
More bullsh*t. Got any proof for that ?
The reason I see it being harder to raise inflation now to 4% is because we can’t lower nominal interest rates.
Last time I checked, the printing presses were operational.
But I don’t think it is as straight forward as thinking we can set a nominal target and markets will magically take care of the rest.
The fact that you are too dense to grasp how expectations work doesn’t falsify them.
From a rational expectations perspective, lowering the real rate of interest pulls future consumption forward.
Oh look, even more bullsh*t.
even at a nominal federal funds rate of basically zero, there isn’t enough consumption
Last time I checked, low interest rates mean MONEY HAS BEEN TIGHT. So yeah, of course you’d have very little consumption !
businesses are refraining from large new investment opportunities […] because their isn’t enough future demand expected to incentivize these new investment.
So you’re saying money is too tight ? I agree.
For there to be adequate demand, households need to spend a larger share of their income, not just spend more in nominal terms.
You really don’t know what aggregate demand means, do you ?
For households to significantly increase their consumption relative to their income, their *real* earnings have to rise.
Nope, you definitely don’t know what aggregate demand means.
And the incentive to invest will be dependent on how much *subsequent* future productive investment opportunity there is because aggregate spending created by aggregate investment in one period will fall short of costs (because of saving) unless there is more *productive* investment going forward.
http://en.wikipedia.org/wiki/Word_salad
lowering the real rate further is unlikely to spur a higher rate of consumption
You really need to stop with this “lowering the real rate” bullsh*t. Print money DOES NOT LOWER THE REAL RATE. Period. End of story. Just stop it.
a higher percentage of national income going to people that save a larger proportion of their income (wealthy people) also depresses demand.
It’s already established you have no idea what aggregate demand means, you don’t need to provide further proof.
I bet you think you’re real clever, but all you do is to hold up your gut feelings as being self-evident. Rationality is hard, bro. But you should definitely try it out someday.
19. October 2014 at 12:20
Regarding the “expectations” you so flippantly dismiss, Nick Rowe has something smart to say
http://worthwhile.typepad.com/worthwhile_canadian_initi/2014/10/inflation-derps-are-people-from-the-concrete-steppes.html
You might want to edumacate yourself on the issue before spewing further drivel.
19. October 2014 at 13:53
Scott, thank you for replying to my response. You are correct about the notion I spoke of regarding periods of nationwide unemployment being “helpful.” My comment had an austrian slant that, in retrospect, I don’t really believe.
You said, “It’s certainly not true that easy money shifts spending from investment to consumption, indeed just the opposite. Monetary stimulus tends to increase investment more than consumption.”
If I said that, I misspoke. I never intended to say that easy money shifts spending from investment to consumption. I do know that monetary stimulus tends to increase investment. I meant to say that from a rational expectations (New Keynesian IS-curve) perspective, a lower real rate shifts future consumption forward to the present. It doesn’t create new demand.
Investment generates demand, but in an amount that is less than the total cost of the investment because people save part of their income. So unless there is an expectation of future demand (not just current demand) sufficient to justify ongoing future investment (not just current investment), then it doesn’t make sense to invest in new production now on a large scale. Yes, Keynes.
If six years of a negative real rate isn’t enough to raise demand yet, there is a good chance that future demand has dried up. So if I were “the markets” I would think a large increase in investment now would not be profitable because there is not enough future demand to support ongoing investment (in dollar amounts that exceed the dollar amount of savings) from the income generated by these investment.
I don’t think that rational expectations are necessarily the truth. But I think they are helpful in thinking about the relationship between current investment and future demand.
You teach economics so you likely have understand these models better than I do, and you may disagree that future demand has dwindled. And you may be correct.
You said, “Don’t even think about interest rates when thinking about investment (never reason from a price change.) Interest rates are not monetary policy, NGDP growth is.”
I do think there is something more fundamental when the Fed funds rate has been essentially zero for six years. It suggests to me that future (planned) consumption has dwindled. And without adequate future demand, there is not an incentive to invest now on a large scale.
Because exceptionally low or negative borrowing rates can have ill effects (such as chaotic capital flows in and out of emerging economies, and possible asset bubbles), I think the zero lower bound is a very different animal.
Say for a moment that my model is correct (not that it necessarily is) that there is not enough future demand to support future investment in excess of savings. Say the reason for this is because new business ventures (such as What’s App?) just don’t require the kind of capital investment that new ventures did in the past (this is a Larry Summers argument). And say this trend (less money needed for new ventures) is continuing to trend down. Furthermore, say that domestic and international savings of U.S. dollars are binding, meaning people would save a larger proportion of their income if they could.
Now say the Fed targets NDGP. But people continue to try to save more as a percentage GDP, and new productive investment outlays trend down. But the Fed is still able to push a 5% growth rate in NGDP which drives the real rate more negative. If this somehow brings people back into the labor force, we achieve full employment, real wages start to rise, demand increases, and the Wiksellian rate goes up, then great! It worked and you’re a hero.
If, on the other hand, people keep trying to save a larger portion of their income than what can be productively invested, then either: 1) it just doesn’t work, the Fed fails in its attempt to target NGDP growth; or 2) there is a large increase in unproductive investment that is only possible because of negative real rates, and ultimately this unproductive investment will lower real growth, or lead to some sort of calamity either here or overseas.
Sorry for the long post. I do acknowledge that my model is not necessarily correct.
19. October 2014 at 14:28
Daniel,
Wow! I’ve really stirred you up. I’m sorry if I sound smug, it’s one of those things that when you write, you come off differently than you try to. That said, you come off as a bully, but I doubt you really are one in real life.
I read Nick Rowe’s post last night and even commented on it. Here’s my comment:
(Posted by: Mark | October 19, 2014 at 01:41 AM)
“Nice post Nick. I makes your point very well. But if it were completely true, then lowering interest rates in normal times wouldn’t spur demand either, but it does. And your post doesn’t explain why inflation expectations have dropped below target. But it is still a great post!”
Regarding your repeated jabs at what I said, I think we are just thinking of different models. (BTW, I enjoyed the one that was a Wikipedia link to “word salad.” It was probably apropos to.)
I think even Scott would agree that you can view NGDP targeting though the lens of real and nominal interest rates, though he thinks it is more helpful to look past that and focus on nominal spending.
If you look closely at my model, through the lens in which I frame it, you’ll see that it is coherent (which doesn’t mean it is correct).
So pulling consumption forward in the New Keynesian IS-curve for example doesn’t really mean consumers see the interest rate go down and borrow money to use to consume with. I view it as investment goes up now which stimulates demand. But it suggests that this investment would have happened at some point anyway, if you have a slump and investment goes down, the presumption is that it will be made up for in the future. By lowering the real rate during that slump, you bring that make up investment forward and it smooths out the business cycle.
So my model is misleading to say the least because it frames things in terms that aren’t necessarily how things work (such as “pulling consumption forward”). Another way for me to frame it would have been to say lowering the real rate pulls future investment forward and spurs consumption. And my point is that 6 years of a negative natural rate without a broad recovery in investment suggests to me that there is something more fundamental than monetary policy going on. And that raising inflation would have the *effect* of lowering the real rate. But my point is that lowering the real rate wouldn’t address whatever is causing the natural rate to drop below zero.
Anyway, I’m not sure I want to thank you for your prickly comments… Oh what hell: Thank you for your comments!
Believe it or not I do find them helpful.
19. October 2014 at 20:32
Daniel, one of your comments directed at me makes a good point (though I didn’t care for the tone of your delivery). I threw around the terms “consumption”, “investment”, and “demand” without being clear. Of course I mean aggregate demand which consists of consumption and investment (as well as government spending and net exports). But what I’m saying is that if firms (on aggregate) don’t believe the demand for their production will be high enough, then they won’t invest in new production. To be clear: new investment raises demand, but it might not raise it enough to pay for the new investment (even with lower financing costs). I’m essentially just restating Keynes’ argument against Say’s Law.
19. October 2014 at 23:54
there is a good chance that future demand has dried up.
Well, like I said. You obviously have NO IDEA at all what “aggregate demand” means, yet you have firm opinions on macroeconomics.
This is you
http://en.wikipedia.org/wiki/Dunning%E2%80%93Kruger_effect
And no, “aggregate demand” is not some sort of magical well which dries up because the confidence fairy says so.
21. October 2014 at 04:37
Mark, You said:
“If, on the other hand, people keep trying to save a larger portion of their income than what can be productively invested, then either: 1) it just doesn’t work, the Fed fails in its attempt to target NGDP growth; or 2) there is a large increase in unproductive investment that is only possible because of negative real rates, and ultimately this unproductive investment will lower real growth, or lead to some sort of calamity either here or overseas.”
It would not fail to boost NGDP if done properly, and if real rates are negative so be it, that would not lead to unproductive investments, as long as rates are set by the market. Where is the market failure in your model?
21. October 2014 at 12:10
Mark –
As Scott said, maybe the biggest lesson in MM is don’t confuse interest rates with the stance of monetary policy. Keep in mind the interest rate an investor demands is partly dependent on his expectations for inflation, which is why interest rates have steadily declined since the Volcker regime began altering the inflation expectations of investors. That’s why a nation can have low interest rates and a more-tight-than-ideal monetary policy, especially in the long run.
The oil shock example above is a great one, because like the curvature of light around a star does to Newtonian mechanics, it exposes the obvious flaw in inflation targeting — no one really believes you should tighten during an oil shock! Price stability is not the be-all, end-all of monetary policy and everyone already knows this. NGDPLT provides a theory of policy that works for oil shocks as well as normal times.
21. October 2014 at 12:22
On the income inequality side (which is controversial, I know), a higher percentage of national income going to people that save a larger proportion of their income (wealthy people) also depresses demand.
This is where people get into a lot of trouble with counterfactuals — depresses demand compared to what? Does a world without Google, Yahoo, and Facebook increase demand because the “share” of income going to lower income quartiles is higher without them? Of course not, we’d all be worse off without them — the whole notion of a “share” of income is facetious. In a free market where people voluntarily decide where to spend their money, income and wealth follow the creation of value. What you’re really talking about is “productivity inequality” — globalization and technology have made it possible for people to create value on unprecedented scales.
21. October 2014 at 13:31
[…] been sent only the second piece, Scott Sumner refers to JB […]