Recent Posts

Wednesday, January 11, 2017


Even if one is not a fan of President-elect Trump, one must be impressed how he managed to shift expectations to end secular stagnation. With the Zero Lower Bound (ZLB) now dead as a door-nail, it is going to be difficult for mainstream economists to make it look like Dynamic Stochastic General Equilibrium (DSGE) models have something useful to tell us about the real world.

This article first discusses some of the academic theoretical issues around the ZLB, and then jumps to the latest economic squabbling based on Paul Krugman's recent comments.

Rogoff Scoffs at the ZLB

Professor Ken Rogoff was a discussant on a ZLB paper at the ASSA conference in Chicago. His comments are described by Professor J.W. Mason in this article.

I will refer the reader to J.W. Mason's article to see his transcription of Rogoff's comments. I just wanted to highlight Mason's remarks at the end of his article.
One way of looking at this is that the ZLB is a device to allow economists like Krugman and Gourinchas and Rey — who whatever their scholarly training, are aware of the concrete reality around them — to make Keynesian arguments without forfeiting their academic respectability. You can understand why someone like Rogoff sees that as cheating. We’ve spent decades teaching that the fundamental constraint on the economy is the real endowment of resources and technology; that saving boosts growth; that trade is always win-win; that money and finance matter only in the short run (and the short run is tolerably short). The practical problem of negative policy rates doesn’t let you forget all of that.
Which, if you turn it around, perhaps reflects well on the ZLB crowd. Maybe they want to forget all that? Maybe, you could say, they take the zero lower bound seriously because they don’t take it literally. That is, they treat it as a hard constraint precisely because they are aware that it is only a stand-in for a deeper reality.
If I wanted to be charitable, I think this is the nicest interpretation of what has been going on in DSGE macro. DSGE macro is an intellectual failure, but we can pretend to get useful results if we stick in silly things like an assumption that the policy rate being allegedly stuck at 0% forever completely changes how the economy functions.

Paul Krugman Triggered Me!

 Paul Krugman wrote a recent New York Times op-ed -- "Deficits Matter Again" -- which has raised a lot of eyebrows amongst supporters of Modern Monetary Theory (MMT), or post-Keynesians more generally. I will discuss my take on his comments further below; the summary is that I think his comments are defensible from a mainstream economics viewpoint, but his hand-waving explanation of the dynamics is either terribly written or outright incorrect.

My gut feeling is that he phrased his article in such a way so as to deliberately drive post-Keynesians crazy. I may be just taking the bait, but I just wanted to throw in some comments.
On the other side, there this article by Gerard MacDonell -- "What did Krugman Know, and When Did He Know It?" In it, he discusses the political context of Krugman's article; in earlier article, Gerard made the same point about "full employment" that Krugman did. As I discussed in an article I published in December, Gerard is using "full employment" in a way that many (wide definition) Keynesians (such as Bill Mitchell) would object to. Although I would phrase things differently, my reading of the current cyclical conditions may not be fundamentally that different than his.

Crowding Out - Sigh

The most objectionable passage (for post-Keynesians) in Krugman's article was the following paragraph.
No, they’re going to blow up the deficit mainly by cutting taxes on the wealthy. And that won’t do anything significant to boost the economy or create jobs. In fact, by crowding out investment it will somewhat reduce long-term economic growth. [emphasis mine] Meanwhile, it will make the rich richer, even as cuts in social spending make the poor poorer and undermine security for the middle class. But that, of course, is the intention.
The highlighted sentence looks to have been engineered to put post-Keynesians in a frenzy -- or it may have been the result of an economically illiterate editor at the New York Times. It is largely irrelevant to Krugman's big idea -- that the Republicans are being hypocritical about their deficit fears.

The phrase "crowding out" is by itself fairly meaningless; historically it was associated with the Loanable Funds fallacy. Is Paul Krugman reverting back to loanable funds thinking?

Perhaps he is, but we should have no reason to care what his particular thought processes are. He is a standard bearer for mainstream economics, and we should look at what that theory says.

If we take the pretensions of mainstream economics at face value, we should be discussing this topic as follows.
  • We start out with the current policy stance as the base line. The economy is at an "equilibrium state," characterised by various levels of interest rates, investment, consumption, and so forth.
  • We can then imagine the policy stance changing (courtesy of President-elect Trump), and the economy snaps to a new "equilibrium" state, with new levels of interest rates and so forth.
  • Paul Krugman's argues that the new equilibrium features higher interest rates, higher consumption, and somehow (?) lower investment.
I can see how one could attempt to describe this process as tax cuts "crowding out" private investment; no appeal to "loanable funds" is required. The problem with the logic is that there is no reason to believe that investment will be lower in the second equilibrium. We commonly hear about excess capacity in practically every sector of the economy; increased demand could easily bring forth greater supply.

Even if we put that debate aside, I am not particularly impressed with the argument.

Firstly, if a tax cut does not do "anything significant to boost the economy," why should we object? The objective of taxation is to regulate demand (to preserve the value of the currency); why not keep taxes at the lowest possible rate to achieve that objective? Presumably, the tax cuts are part of the starve the beast" political strategy, which Jared Bernstein discusses in his article. This would obviously be of concern to Democrats, but the strategy of invoking bond market vigilantes to counter it has been a strategy with a multi-decade record of failure.

Secondly, why should we care if interest rates are higher (unless one is a leveraged long)? We move away from the dreaded ZLB, and we reduce the pressure on the retirement savings complex, which is a key point of weakness for the middle class. Unless one believes in dubious theories about "fiscal space," the objective of policy should be to raise interest rates.

I would note that I have no fear whatsoever about the inflation outlook. A large rise in inflation would be the risk of a massive shift in fiscal policy. However, I have serious doubts that policy settings would ever be changed that radically, no matter how exuberant the Republicans get. Core inflation has been anchored for decades, and a few timid tax cuts are not going to change the situation.

Concluding Remarks

Secular stagnation and the ZLB are dead. DSGE macro will need a new gimmick in order to pretend to be relevant. On the policy front, we really need to wait to see actual policy shifts in order to be able to have an idea whether they matter.

Appendix: The Politics of Deficit Phobia

In Gerard's article, he made some comments about a blogger friend who I assume is myself. He objected to my amusement at "Democratic-leaning" economists switching their tunes on fiscal policy once Trump was elected (similar to how some Republican leaders have suddenly become much less worried about government debt post-election).

I generally avoid writing about politics, but I do not claim to have "non-partisan objectivity." Partly my reasoning is commercial: I am writing primers about bond market economics, and see little reason to pointlessly anger potential readers. However, it reflects the fact that I am one of the few Canadian Prairie Populists outside of an old folks' home.

I doubt that many of my readers are aware of the Prairie Populist world view, but it safe to say that both Hillary Clinton and Donald Trump are not exactly the types of politicians that are approved of. Correspondingly, my comments about their partisans should not be expected to have an approving tone.

(c) Brian Romanchuk 2017


  1. Nice post. I agree with you that what Paul Krugman wanted to do was point out that Republicans were being hypocritical about deficits once again. But that is a fairly easy argument to make even without referencing loanable funds based "evidence" to support it.

    I thought Mitchell's criticism was fairly comprehensive, but you have managed to come up with an alternative way to interpret how Krugman supports his argument that would not be based on loanable funds theory. It is imaginative and demonstrates your creativity. But like you say, it is not much of an argument. And I really doubt he was thinking along those lines.

    I had managed to leave Mr. MacDonell a comment recommending Mitchell's blog post. According to his update on his post, this prompted him to look up Ken Rogoff instead. (Although he did post a polite reply with a nice comment about Brian). It is not apparent that he consulted the Mitchell blog, but it is very clear that he does not wish to discuss MMT based criticisms. Oh well, what can you say?

    1. Apparently Mr. MacDonnell has decided to discuss MMT. He has a much more reasonable contribution than the recent one from Richard Holden, but he is still somewhat wrong. He also has revised his post about Krugman's argument- "What did Krugman Know and When Did He Know About It?" So my above comment is not entirely relevant at this point.

      The latest post about MMT is at

    2. Thanks.

      I read the article, and started thinking about a response. Probably this weekend's article.

    3. Well if you do have a response post, you might consider Bill Mitchell's latest answers to his weekly quiz, especially to questions 2 and 3.

      I have been searching for years now for good criticisms of MMT. For things that show it does not describe the system accurately. Because it in all honesty sounded too good to be true. But I have not found them.

      It seems to me that if someone could demonstrate that MMT was wrong they should show why Bill's answers to question 2 and 3 are wrong. MMT is a very logical theory built in a large part upon these foundations. I have been waiting for years for someone in the banking business or bonds to say No- it doesn't work that way and provide a decent answer as to why. Still waiting and until and if it happens MMT is the best description I have heard.

    4. One other thing you might consider is my little rant here

      and more importantly, Bill Mitchell's reply. Anyways, I think I outlined some reasonable avenues to refute MMT there.

  2. "We commonly hear about excess capacity in practically every sector of the economy; increased demand could easily bring forth greater supply."

    I'd be very interested in seeing you develop that thought further.

    1. It's an assertion on my part, based on my understanding of Verdoorn's Law and my guesstimation about "real" capacity utilisation. It's also possibly sloppy writing; I should have written investment and supply (since investment was what I had in mind). Although I will admit that I am waving my hands, it's probably less hand-wavy than asserting that investment will fall...

      (The investment *share* might fall if consumption rises, but that's just being a smaller slice of a bigger pie. Not something I see as a problem.)

      I was planning on doing some more in depth business cycle work, which I will mix in with rudimentary SFC modelling. That should expand on my guesswork here (hopefully not with the opposite conclusions). Just a last cycle of edits, and no more analysis of money for me...

  3. "We commonly hear about excess capacity in practically every sector of the economy; increased demand could easily bring forth greater supply."

    Not from J.L.Y.?

    "The economy is operating relatively close to full employment at this point," she said, "so in contrast to where the economy was after the financial crisis when a large demand boost was needed to lower unemployment, we're no longer in that state." '

  4. Brian,

    Regarding your description as one of the few Canadian Prairie Populists outside of an old folks' home, perhaps you are an old soul?

    I don't have time right now to read the links, although I already took a look at J.W. Mason's article the other day. Perhaps I missed your explicit reasoning but I don't see any case made for this main idea: Zero Lower Bound (ZLB) now dead as a door-nail, it is going to be difficult for mainstream economists to make it look like Dynamic Stochastic General Equilibrium (DSGE) models have something useful to tell us about the real world.

    I assume Fed is looking at some measure of the US output gap to decide whether a fiscal and/or monetary stimulus might generate significant inflation. I see some charts showing the US output gap is nearly closed. I don't think the US economy is anywhere near full capacity or full employment with regard to putting upward pressure on wages or on interest rates but it is just intuition and not based on a model. How do developing conditions now tend to RIP ZLB and invalidate DSGE models?

    I apply a simple model of tax cuts on the wealthy. If middle class cannot go into debt to purchase all the output then tax cut to the wealthy will not stimulate further demand, generate inflation, or increase interest rates, since the wealthy have no profit incentive to expand capacity in anticipation of future demand. If King Kong forces the Republicans to stimulate the economy via tax cuts to the middle class or deficit spending for infrastructure or other stimulative govt programs then it may send a signal to expand productive capacity. Or if middle class balance sheets are clean enough to take up more private debt, and such debt is authorized by financial dealers, it will stimulate more productive capacity to come online. This is all abstract logic based on the psychology of investors and cash flow generation and is of course not tied to a model of the economy at this time.

    1. DSGE models were only generating useful stories because of the imposition of the zero bound. With rates back away fromzero, they are back to being largely useless again. Very little to do with current cyclical conditions.

    2. Who or what mechanism imposed the ZLB? How does ZLB tend to validate DSGE for some observers or true believers? How do recent events cause one to regard DSGE as invalid based on what is perceived as a ZLB?

      I tend to view the US financial system as a dealer network, which is a model discussed by Perry Mehrling, with Fed and Treasury as the financial dealer(s) and investor(s) of last resort. If one wants to experiment with credit dealer systems having no dealers and investors of last resort then ZLB is not a boundary at all but rather a point on a continuum of positive or negative interest rates that could manifest. The concept of negative interest rates is nothing special in a dealer network model: if the credit system is forced to contract balance sheets it means financial dealers are selling assets and destroying liabilities. Banks would impose fees at a faster rate then they are making loans and investments, and the result is negative interest rates, reduction of the quantity of debt and equity in the financial sector. I do not wish to study DSGE models in depth but I am interested in why economists regard such models as valid or invalid based on evidence.

    3. Why do DSGE models are "validated" by the ZLB? That's what is discussed in J.W. Mason's article (and in Rogoff's comments). In the absence of the ZLB, DSGE macro implies silly things, such as fiscal policy having no effect. Since the whole area is a bunch of hand-waving that pretends to have a lot of mathematics in it, I don't really can't say anything about it, one way or another.

    4. I appreciate your responses. According to J.W. Mason it seems some economists use the concept of ZLB to make up stories that are coherent with their school of economic thought even if the boundary can be seen as nothing special by a Keynesian.

      I would reverse the logic of fiscal policy effects: if quantitative easing and/or fiscal policy does not prevent negative interest rates then it cannot be effective at stimulating economic growth. Expected future profits should be lower in an aggregate economy with negative interest rates since aggregate financial savings would be getting smaller over time. I could still imagine a DSGE model that appears valid under somewhat realistic conditions, however, it should not predict that negative interest rates accompany or cause economic growth.

  5. One of the dangers facing DSGE models is discussed here by Steve Keen.

    I found it quite humorous.

    1. well I really am no good at posting links. That one is wrong.
      I will try again.


  6. Steve Keen's sarcastic paper refers to this 26 page paper by Paul Romer which explains the trouble with DSGE models macroeconomics:

    Romer uses the so-called Volcker deflation circa 1980 as the event which tends to show the invalid theory and application of RBC (real business cycle) and DSGE (dynamic stochastic general equilibrium) models. He says the models contain imaginary shocks from mysterious outside forces, too many input parameters that are adjusted without a basis in reality, no real basis in microeconomic foundations, among other criticisms. The assumptions made by "experts" are called Facts With Unknown Truth Value (FWUTV) on the bottom of nominal page 6. I think he says Keynesian models suffer some of the same problems but are more honest about expert assumptions and the limits of the model.

  7. Krugman unequivocally said the Loanable Funds was "true" 2014

    (Video clip)


Note: Posts are manually moderated, with a varying delay. Some disappear.

The comment section here is largely dead. My Substack or Twitter are better places to have a conversation.

Given that this is largely a backup way to reach me, I am going to reject posts that annoy me. Please post lengthy essays elsewhere.