Nick Rowe’s wisdom, New Keynesianism vs. NeoFisherism, and Fed incompetence, all explained in one 7 minute bicycle video

Here it is.

HT:  Tyler Cowen



39 Responses to “Nick Rowe’s wisdom, New Keynesianism vs. NeoFisherism, and Fed incompetence, all explained in one 7 minute bicycle video”

  1. Gravatar of Jean Jean
    22. August 2015 at 07:22

    That was great – thanks for sharing!

  2. Gravatar of Ray Lopez Ray Lopez
    22. August 2015 at 08:04

    Absurd video, a waste of seven minutes. All he was showing is ‘muscle memory’, nothing else. If you stop practicing a sport, like shooting a basketball, shooting pool, or even playing chess, the same thing happens for a while until your old form comes back.

    Sumner: please reprogram yourself to believe in the data and quit trying to find patterns where none exist. Money is neutral.

  3. Gravatar of Jose Romeu Robazzi Jose Romeu Robazzi
    22. August 2015 at 09:22

    Great video, it looks like it is hard to change a paradigm aproach, but it turns out that when it happens, it is intantaneous !!

  4. Gravatar of Nick Rowe Nick Rowe
    22. August 2015 at 09:59

    At the Ottawa exposition, about 20 years ago, they actually had some reverse bikes like that. If you could ride one for 20 feet, you won a prize. I tried it, and failed hopelessly.

    Here’s an inverted pendulum

  5. Gravatar of Tom Brown Tom Brown
    22. August 2015 at 12:01

    Thanks Scott, nice video.

    To complement Nick’s inverted pendulum video, I’m a fan of this triple inverted “self-erecting” pendulum. Enjoy.

  6. Gravatar of Major.Freedom Major.Freedom
    22. August 2015 at 12:31

    Ray, why do you keep lying to yourself?

    Money is not neutral EVEN BY YOUR OFFERED STANDARD.

    A rate between 3.2% and 13.2%, which means an average of about 8%, is not zero.

    Not de facto.

    Not effectively.

    Not basically.

    Not essentially.

    Do you know what is a waste of time? Reading your fallacious posts over and over again.

    You should reprogram yourself. There is a bug in your processing code.

  7. Gravatar of ssumner ssumner
    22. August 2015 at 12:44

    Nick and Tom, Put that machine in charge of the Fed!

  8. Gravatar of Jason Smith Jason Smith
    22. August 2015 at 13:09

    These (unstable) increasing over-corrections look like they appear in the monetary base data as I show at the bottom of this post:

    However, the mechanism I describe is a bit different. Instead of inputs going the wrong way, the response to a given input falls. This leads to over corrections like in the bike example, but doesn’t necessarily crash the bike.

  9. Gravatar of Major.Freedom Major.Freedom
    22. August 2015 at 13:18

    I knew there was always a problem with the MM go to response to the free market advocates. It has a name!

  10. Gravatar of Tom Brown Tom Brown
    22. August 2015 at 13:49

    Scott, the next two videos that got automatically cued up for me were pretty interesting too: [1], [2].

  11. Gravatar of Jose Romeu Robazzi Jose Romeu Robazzi
    22. August 2015 at 16:18

    That is a two dimensional inverted pendulum, to make it happen one need a kalman filter approach (Rodrick Prescott raise a bell anyone) and computing power to make an 8 x 8 adaptive control matrix … I learned how to do that in college, although I did only simulations of the that. The real thing is much more fun.

    The interesting thing about this is that a few years back, I thought about using the same adaptive control ideas in order to model interest rates. The short term is the base of the pendulum, the long term (the one the fed actually wants to influence) is the loose end of the pendulum. Monetary authority can control short rates. When you are very far off your objectives, large control action is needed, but once one is in equilibrium, very minor control is needed in order to keep the loose end up straight.

    But now I think forget interest rates, just do NGDP futures, targeting the forecast. If Prof. Sumner here is right, once the monetary authority is very close to target, very minor adjustments are needed to keep the pendulum up… if one uses the correct meter (NGDP futures) to see where the loose end of the pendulum is …

  12. Gravatar of E. Harding E. Harding
    22. August 2015 at 18:57

    The video has absolutely nothing to do with the title.

  13. Gravatar of Tom Brown Tom Brown
    22. August 2015 at 20:41

    Scott, what is the “New Keynesianism vs NeoFisherism” part of the video? Specifically which bit is “New Keynesian” and which part “NeoFisher?”

    Also, I’ve seen something similar to that video but with an aircraft. When I was in HS I worked at the airport in exchange for sailplane (glider) lessons. My instructor took his sailplane in for “servicing” and the morons reattached the rudder cables backwards… so when he pressed the right one, it went left and vice versa. He didn’t make it more than about 20 yards down the runway (behind the tow plane) before he veered violently off into the weeds, snapping the tow rope. Nobody got hurt.

  14. Gravatar of James Alexander James Alexander
    22. August 2015 at 23:23

    Personally, it helped me with Information Transfer Economics too. Much clearer …

  15. Gravatar of ssumner ssumner
    23. August 2015 at 04:40

    Jason, I don’t understand your claim that M0 never grows by more than 15%. M0 generally refers to the base, which has recently grown much faster.

    Jose, Good point.

    Tom, Each side accuses the other of this fallacy. Most economists would say that the NeoFisherians have it wrong. The bike analogy is that if you want lower interest rates, you first have to raise them. In 2011 the ECB raised interest rates, and that drove them even lower than in the US.

  16. Gravatar of Patrick R. Sullivan Patrick R. Sullivan
    23. August 2015 at 06:37

    ‘Put that machine in charge of the Fed!’

    Scott, Do you know any Google nerds? Now there’s real power;

  17. Gravatar of Ray Lopez Ray Lopez
    23. August 2015 at 06:37

    @MF: “Money is not neutral EVEN BY YOUR OFFERED STANDARD. A rate between 3.2% and 13.2%, which means an average of about 8%, is not zero.”

    No, that’s not what a confidence interval to 95% confidence level means (from Bernanke et al’s 2003 paper). It means it’s equally likely you get 3.2% effect as you would get 13.2% effect. You don’t ‘average’ the two endpoints to get a ‘most likely’ number, that’s not how confidence intervals work.

    Let’s face it: 3.2% or 13.2% out of 100% is trivial. That’s the effect of monetary policy says Bernanke et al (2003). No wonder Bernanke later said, in testimony to Congress, that people expect too much from the Fed. He was probably thinking of this paper he wrote years before. A paper BTW Sumner has probably never read (afraid he might learn the truth, like a naive Roman Catholic who is forbidden to read the Bible and afraid to do so). Sumner, you can’t handle the truth.

  18. Gravatar of mbka mbka
    23. August 2015 at 06:51

    Scott, I assume that you’re the guy who learned to ride the “other” bike of NGDPLT … now all sorts of things like inflation targeting or reasoning from interest or price changes seem absurd to you 🙂

    About the other comments on overcorrecting etc. I don’t think that is the issue here. Problems of overcorrecting occur easily in ordinary control systems due to phase shift. Basically, the control input comes after the sensor input of a problem and therefore always a little too late. At a specific time, the control input is exactly out of phase with the initial movement and will accidentally reinforce it. (simplistically explained). This happens in many systems – from suspensions to electronics, or formally, say, in the discrete logistic equation. You will also know it if you ever tried to ride a sidecar motorcycle – one that is steered the ordinary way. But due to the drag on the sidecar side one keeps overcorrecting at first and the bike will often shake out of control. In all these systems, the problem is that the solution makes the problem worse and violent oscillations occur. Solution: either add friction (the shock absorber in mechanical systems), or slow the response down below the original system’s danger zone, or make it very fast with safeguards against oscillations at the upper frequency end (this is used in electronic circuit control). BTW in the sidecar case, the solution is Zen: just relax … give up the attempt at control … let the steering shake a little… it will stay approximately stable.

    Scott’s NGDPLT, controlled by prediction market, is a more elegant application of the systems control problem because it is a “feedback” before the problem even occurs and the phase lag problem, ideally, might disappear. This solution is analogous to another electronic solution, feed forward instead of the usual (delayed) feedback.

    I always wondered why finance people don’t seem to talk much to systems control theory people because the problems are analogous.

  19. Gravatar of ssumner ssumner
    23. August 2015 at 14:36

    Patrick, If only I controlled Google.

    mbka, Very interesting comment.

  20. Gravatar of Ray Lopez Ray Lopez
    23. August 2015 at 17:15

    OT: on wage non-stickiness, and why it does not even matter if wages are slightly sticky. John Robertson of the Atlanta Fed: “As noted by Elsby, Shin, and Solon (2014), it is not clear that an increase in wage rigidity would constrain the hiring of new workers more than it constrains the retention of existing workers.” – has Sumner read this paper? No. Add it to the other papers he’s not read: Bernanke et al (2003) showing the Fed has little influence over the economy, and, “Did the Reserve Requirement Increases of 1936-37 [NO] HAELIM PARK and PATRICK VAN HORN Article first published online: 28 JUL 2015.

    Note that the graphic in ‘Economist’s View’ for Friday, August 21, 2015, “Evidence for Sticky Wages” shows a full 32% reported a DECLINE in weekly wages from a year ago. So much for the canard that wages don’t decline.

  21. Gravatar of Christian List Christian List
    23. August 2015 at 18:14

    @Ray Lopez
    You mentioned this paper by Bernanke et al (2013). What is the name of the paper again?

  22. Gravatar of TallDave TallDave
    23. August 2015 at 19:45

    Not to mention Xi Jinping.

  23. Gravatar of TallDave TallDave
    23. August 2015 at 19:49

    mbka — Yes, expectations are the D in the PID controller of economics.

  24. Gravatar of Steve Steve
    23. August 2015 at 20:49

    The “control theory” mbka references is best illustrated by the example of the Tacoma Narrows Bridge, a.k.a., Galloping Gertie
    (ff to 3:00 if you don’t want to watch the whole thing)

    This is what happens if a force is applied at the underlying harmonic frequency of the structure.

    I suppose it is possible that a bad monetary reaction function could create a Gallopin’ Gertie economy, but I think it is bigger risk with cycles of bank regulation, trade openness/protection, or commodity investment.

  25. Gravatar of Tom Brown Tom Brown
    23. August 2015 at 22:58

    Scott, Jason uses MB for the base and M0 for the base minus reserves (i.e. cash in circulation).

  26. Gravatar of Dan W. Dan W.
    24. August 2015 at 04:40


    Is not this article in Vox validation of the theory of “bubbles”? It is a fair argument to say this is a human caused failure. But where did the Chinese regulators err? Was it in enabling the buying of stocks on margin or in restricting it?

    Between June 2014 and June 2015, China’s Shanghai Composite index rose by 150 percent. A big reason for the stock market rally was that a lot of ordinary Chinese people began investing in the stock market for the first time. More than 40 million new stock accounts were opened between June 2014 and May 2015.

    And many have been buying stocks with borrowed money. The Chinese government used to strictly limit this practice, but over the last five years the government gradually relaxed those regulations.

    Earlier this year, the authorities became concerned that the stock market’s rise had become unsustainable. So they began to tighten limits on debt-financed stock market speculation. The stock market peaked in June and then began to fall quickly. That caused regulators to change their minds again. In early July, they made aggressive efforts to push stock prices back up.

  27. Gravatar of Ray Lopez Ray Lopez
    24. August 2015 at 06:31

    @Christian List – here is the paper. If you’re a monetarist, read it and weep.

    Measuring the Effects of Monetary Policy: A Factor-Augmented Vector Autoregressive (FAVAR) Approach * Ben S. Bernanke et al (2003) “Apart from the interest rates and the exchange rate, the contribution of the policy shock is between 3.2% and 13.2%. This suggests a relatively small but still non-trivial effect of the monetary policy shock. In particular, the policy shock explains 13.2%, 12.9% and 12.6% of capacity utilization, new orders and unemployment respectively, and 7.6% of industrial production”

  28. Gravatar of Vivian Darkbloom Vivian Darkbloom
    24. August 2015 at 07:17

    @Dan W

    I pointed out the link to this chart (also produced by Tim Lee of Vox) on July 8, in response to Scott’s post entitled “The Chinese Stock Market Crash”:

    This might not be proof of “bubbles”, but it may add credence to the idea that if you believe in the wisdom efficient markets, you’d better make sure you are following an efficient market, or even a market. What does it take to constitute a “market”, really, much less an “efficient” one?

  29. Gravatar of TravisV TravisV
    24. August 2015 at 08:50

    Bob Murphy:

    “Scott Sumner, Vindicated”

  30. Gravatar of Ray Lopez Ray Lopez
    24. August 2015 at 09:34

    @TravisV – a confusing post by Bob Murphy, I *think* he was being sarcastic in crediting Sumner, who predicted no China crash. It’s hard to tell however. Murphy writes like an economist–too clever by half, esoteric, and two-faced. Recall Truman’s “one-armed economist”, with all kinds of caveats. That was true over sixty years ago and is true today.

  31. Gravatar of Tom Brown Tom Brown
    24. August 2015 at 09:51

    O/T Scott, I saw this in a recent post by John Cochrane:

    “Financial crises are always and everywhere about debt, especially short term debt. Lending more, encouraging more bank leverage, reducing reserves and margin requirements, means that when the downturn comes a needless wave of runs and defaults follows.”

    “Inevitably, it seems, another downturn will come, another set of books will have been found to have been cooked, and then we will find out who lent too much money to whom. US investment banks, 2008, strike one. Greece, 2010, strike 2. China, 2015, strike 3? Do we no longer bother closing the barn doors even after the horse leaves?”

    What’s your level of agreement with that? Say from 0% to 100%. Also, how representative is that of the Chicago School outlook?

  32. Gravatar of Tom Brown Tom Brown
    24. August 2015 at 09:53

    … BTW, when I was reading that from John, for a second there I thought I was reading Steve Keen or maybe even William K. Black.

  33. Gravatar of TravisV TravisV
    24. August 2015 at 11:09

    The past couple days, the dollar has weakened dramatically, which I find rather confusing……

  34. Gravatar of ssumner ssumner
    24. August 2015 at 11:55

    Thanks Tom, I think all models predict that currency velocity declines when interest rates fall to zero.

    Dan, So government intervention can impact stock prices? Is this a surprise? I’d add that there is very little evidence that governments can control stock prices, even when they intervene. After all, do you think Chinese stock prices have been behaving in a way that is consistent with the government’s policy goals?

    Ray, You are even slower than I thought. Read my comment over there.

    Tom, It’s about debt in the sense that debt defaults rise sharply. But debt is not the underlying cause, in most cases.

    Travis, See my new post.

  35. Gravatar of Dan W. Dan W.
    24. August 2015 at 16:30


    We can agree that government has little control over the impact of changes in financial regulations. I don’t know if the Chinese government understood how leveraged speculation would take over its stock market. I don’t know if the US government understood the degree to which buyers and lenders would turn the US housing market into a giant ATM and Casino complex.

    Where we disagree is the economic ramifications when debt fueled speculation goes bad. As expressed in your answer to Tom you hold the view that NGDPLT can prevent most catastrophic debt defaults. On one hand I agree. However, I believe that any debt defaults NGDPLT prevents in the short term will be all the greater when the default dominoes do start to fall.

    One of the things that stands out in your defense of NGDPLT is how little care you have for financial engineering. Is this because you don’t think financial structure matters? But of course it matters. I hope we agree that financial structuring enables greater capital investment which enhances economic growth. What happens when that structure crumbles? What happens to economic growth?

    You have an extremely idealistic view of the financial world as it seems you believe that capital can swiftly move from the hands of bad managers to good managers and all will be well. If only that was so. The truth is the process is much more messy as financial liquidation is subject to legal and political wrangling. And the retraining of human capital present another lag. It is this friction in the movement of capital that crushes the dream of NGDPLT.

  36. Gravatar of ssumner ssumner
    25. August 2015 at 05:38

    Dan, As usual you haven’t addressed any of my previous arguments. You said:

    “You have an extremely idealistic view of the financial world”

    I have a very negative view of the US financial system, which is horribly distorted by bad government regulations. I have a fairly positive view of Canada’s financial system, although it is far from perfect.

  37. Gravatar of Dan W. Dan W.
    25. August 2015 at 06:40


    One more thing. NGDPLT presumes there exists a correct trend of nominal economic activity. But this assumption fails to take into account fundamental changes in the economy due to technology and social progress.

    If “cold fusion” was discovered and this decreased the cost of electrical power by 99% would not GDP decrease? The GDP contribution from fossil fuel production and consumption would collapse.

    Or consider if a pill, made of common elements, was discovered that cured 99% of sickness. Would this not cause a decrease in GDP as spending on medical services & drugs would collapse?

    Yet despite this decrease in GDP would humans not be better off? Resources spent on one set of “required” activities could then be spent on activities of individual choosing – I suspect many people would take advantage of the “savings” by working less and enjoying more leisure.

    What happens if NGDPLT assumes a model of economic growth that is artificial or otherwise not accurate? What are the risks of purposely inflating an economy that is enjoying, or wants to enjoy, deflationary progress?

  38. Gravatar of Jose Romeu Robazzi Jose Romeu Robazzi
    25. August 2015 at 11:49

    @ Dan W.

    Assuming your point is that maybe NGDP does not describe the world, and assuming that perhaps wild variation in NGDP is “normal”, I would say the following:

    Taking the equation of exchange MV = PQ (or MV=PT). For various reasons, V changes, and that creates volatility in PT (assuming the monetary authority can control M, which is not unreasonable to assume). Volatility in PT is bad, business suffer, consumers as well, but why should we take PQ volatility as “normal”. When PQ growth rate is constant, that increases predictability, and prevents financial friction. In summary, what people in this blog advocate is change your frame of reference, why should M be the basis for “money” ? Instead, take MV, and try to make MV growth as constant as possible, because the “hot potatoes” effect will lead to PT growth to be as constant as possible. And that is good.

  39. Gravatar of ssumner ssumner
    26. August 2015 at 05:31

    Dan, You said:

    “One more thing. NGDPLT presumes there exists a correct trend of nominal economic activity. But this assumption fails to take into account fundamental changes in the economy due to technology and social progress.”

    You’ve been commenting here for years and still don’t know the difference between real and nominal GDP?

Leave a Reply