In a recent blog post, titled Payrolls and Paradigms, Professor Paul Krugman makes the case that current unemployment figures seem to suggest that the “pessimist” in government, or those who believed that the stimulus should have been larger than what it ultimately was, were correct. Although any reader probably already has a clear idea on the basic direction of my beliefs in this case, the criticism I presented in my response on his blog was probably different than most expect. Enough has been said to disprove the idea that expansionary fiscal policy can positively effect the economy, and so to continue on the subject seems fruitless. Instead, my aim is to bring up the superiority in theory over empiricism. My comment, no doubt, was influenced by last night’s reading of Hayek’s Prices & Production and Other Works: F.A. Hayek on Money, the Business Cycle, and the Gold Standard. I have just started, and although the reading is quite difficult (Hayek’s writing style is not very modern, and seems very Germanic) I have been immediately influenced (his point on theory versus empiricism is clearer than Rothbard’s in America’s Great Depression):
Dear Dr. Krugman,
How to do you pretend to establish exactly how much stimulus an economy needs? Whether one economist suggests spending $1 trillion, and another $2 trillion, both economists ultimately pull these figures out of thin air. There is no logic behind these exact numbers, and one cannot judge the validity of these figures based upon any empirical research (we could go as far as to accept Nobel laureat
e Friedrich Hayek’s belief that no theory could be judged valid based on empirical evidence—see Monetary Theory and the Trade Cycle).
The opinion that current events suggest that the stimulus should be greater seems to be entirely based on the a priori belief that greater government spending leads to higher productivity, at least in the event of a liquidity trap. But, based on the same empirical evidence one could also say that fiscal stimulus is entirely ineffective.
I did not come here to argue against stimulus. Instead, I merely aim to make you aware of the fact that empirical evidence cannot accurately provide the information on what direction to take from this point. When economists, such as yourself, decide to pull out arbitrary figures, largely based on conclusions derived from empiricism, they lose much of their reputation. Fiscal policy cannot be decided by taking statistics and then deciding direction based on beliefs assumed to be correct, but are really a posteriori, because there is obviously no sound theoretical backing. If there was, then the figures presented would be less arbitrary.
I can only conclude by saying that if the debate on the size of the stimulus was really between the two camps you illustrate, then both camps are wrong. If that is the case, then this country is truly doomed.
Jonathan -
A couple questions – first, what makes you think that Krugman is proposing “exactly how much stimulus an economy needs”. These are always estimates. Everyone recognizes they are always estimates that are liable to error. It seems to me that your primary critique is his presumption of exactness, and I would argue that no such presumption exist. When it comes to empirical estimates, I often get the impression that a lot of the Austrian epistemological hand-waving is just a fear of committing to a potentially wrong guess. That doesn’t seem very constructive to me.
Second – you’re overstating the variation in stimulus estimates. My understanding is that most analyses that plugged values into a parameterized model came up with around a trillion to 1.3 trillion dollars. The Obama administration thought that was politically unfeasible so they low-balled it. But I at least haven’t heard anyone come up with an actual estimate to justify the lower figure. So you have the ones centering around a trillion, and then you have the Chicago School, strict Ricardian Equivalence people who “modeled” a $0 stimulus as ideal. But if you really think that among stimulus advocates the estimate was that wide (i.e. – between 1 and 2 trillion), I for one would be interested in seeing some citations for that.
Third – I’m not aware of any empirical evidence that stimulus doesn’t work at the zero lower bound. I am aware of several empirical papers that suggest it is considerably weaker (some suggest non-existant) at higher interest rates, because private investment is crowded out. Granted, I haven’t thoroughly reviewed the empirical literature, but that’s what I’m aware of. So I’m personally curious what exactly you’re talking about when you write: “But, based on the same empirical evidence one could also say that fiscal stimulus is entirely ineffective.” You must have something in mind – what do you have in mind here?
Fourth – regarding: “When economists, such as yourself, decide to pull out arbitrary figures, largely based on conclusions derived from empiricism, they lose much of their reputation” I would ask you why you call these figures “arbitrary”. “Arbitrary” implies determination by chance, whim, or impulse. “Arbitrary” doesn’t seem to pertain to the figures Krugman presents at all – EVEN if you disagree with the figures, I don’t see how you can justify calling them “arbitrary”.
Fifth – regarding: “because there is obviously no sound theoretical backing”. You really need to acquaint yourself with the vast theoretical literature justifying these estimates. A good place to start is here: http://faculty.wcas.northwestern.edu/~yona/research/Multiplier-version12.pdf
Arguing with Krugman is like arguing with a wall, at the end of the day, you get nowhere.
Likewise, one needs not even to scratch the surface to see that Krugman’s thought process has already been vitiated as he is a staunch ally of Democrats and the modern day “liberal”.
In order for one to become economically sound per se, one needs to become an apostate of the state.
Since we know where Krugman stands, it will only take a matter of seconds to dismantle his argument.
Keynesianism has failed in the past and it will fail again.
The “Keynesian Revolution” is nothing more than a devolution and has made society as a whole even more primitive.
But, then again I’m sure with enough tweaking and all his prowess, Dr. Krugman will find the perfect blend!
Keynesianism is as irrelevant to Economics as “Deutsche Physik” was to Physics.
The only problem is the fact that almost every academic institution and think-tank has whole heartedly adopted this sophism as indisputable fact.
Yet we wonder why our standard of living has regressed so quickly.
If you don’t believe that those numbers that were essentially pulled out of a hat aren’t arbitrary, then no one can help you.
It doesn’t require much acumen to see that the direction of the economy can’t be centrally planned by a myriad of economists.
Daniel Kuehn,
I fear that you are focusing too much on specifics, which when changed to your liking have little bearing to the general direction of my post. That said, your first and second questions become irrelevant and misguiding. In my response to Professor Krugman nowhere did I criticize him or any policymaker on a specific figure, as any figure produced would be, in my opinion, wrong. So, with that said, I believe that you have missed the point. The general critique is not against exactness, but against the methods employed by which to decide whether or not we need a larger stimulus package. I though I made it quite clear that the debate is between theory and empiricism, not on what size the stimulus package should be (I even clarified in the original comment that I had not come to debate the size of the stimulus, or whether there should be one at all). Continuing, it matters not what the variation between different estimation of what the stimulus should be. That was hardly my point, and it is obvious that the figures I provided are completely arbitrary. As I said above, they were used to illustrate a point: a point that you apparently missed.
Before continuing, I want to comment on your third point. Your terminology is a bit ambiguous, but it seems as if you are turning this back into a debate on whether or not stimulus works during a liquidity trap. I thought we were having the same discussion elsewhere, where you have yet to respond. I do not want to begin the debate anew, because it feels as if by doing so we are simply arguing in circles, without any actual progress being made (I am obviously placing the blame on you, but the blame could also be mine; obviously, I am biased towards myself).
Finally, the figures are completely arbitrary! Even if they were based on mathematical formulas, which they were not, these formulas themselves would largely be composed of arbitrary constants which are no subject to change (as they should be). Deciding fiscal policy in this matter is inherently inferior. Mathematics is great for modeling and for showing general trends, but is horrible for accuracy (ironic as that may be); this is true even down to the simplest long-run equilibrium models (because long-run equilibrium simply does not exist). The paper you link to is an excellent example of my point.
But, the discussion is already veering off topic. I have no intention of repeating myself, so I can only suggest to re-read my comment.
Re: “Before continuing, I want to comment on your third point. Your terminology is a bit ambiguous, but it seems as if you are turning this back into a debate on whether or not stimulus works during a liquidity trap. I thought we were having the same discussion elsewhere, where you have yet to respond.”
Yes – I couldn’t go on for ever. I started to feel like you were asking questions I had already answered. The liquditiy trap wasn’t the point of my third question except insofar as you raised the issue of the liquidity trap in your letter. The point of my third question was two things you said:
1. That his stimulus recommendations are based on an a priori belief in government spending, and
2. That empirical evidence supports the idea that stimulus is ineffective [presumably you're implicitly meaning "ineffective under current conditions"].
The first point is quite obviously dead wrong. A priori beliefs are beliefs held independent of experience, and Krugman’s beliefs about stimulus are quite obviously based on experience. On your second point I just asked you for a citation. For a declaration like that you must have had something specific in mind when you wrote it.
RE: “The paper you link to is an excellent example of my point. ”
Explain this a little more. The “constants” in that model are allowed to vary, from my understanding.
RE: “Deciding fiscal policy in this matter is inherently inferior.”
But the question is – inferior TO WHAT. Yes, estimates of the size of the stimulus are based on empirical estimates of past downturns. Of course that’s going to be imperfect – that goes without saying. But what is the alternative? What is it inferior to that could provide a better strategy? Krugman’s method combines the conclusions of theory and the conclusions of empirical work that both converge on roughly the same fiscal policy strategy. Of course the empiricism isn’t perfect. It’s “inferior” in that sense. But the fact that it’s not an ideal approach doesn’t make it a bad approach. I knew your economics was Austrian – I didn’t realize your epistemology was too. That’s unfortunate.
Daniel Kuehn,
I think that you still have to re-read the comment that I wrote in the post, because I feel that you are arguing with me without actually knowing the position that I hold. I will respond to the largely tangential points that you present, and then I will reiterate what the point was so that maybe we can get this discussion back on track (although, it doesn’t seem as if it was ever on track).
In his blog post, Paul Krugman said that our current unemployment figures suggest that the stimulus should have been higher. I responded by saying that you cannot make a decision like that based only on the unemployment figures. Empiricism cannot tell us whether or not the unemployment figures are lower than they would have been without stimulus, or if the stimulus was completely inefficient. In fact, there are many theories which could fit to the statistics. The validity of a theory cannot be made through empiricism, only through logic. In his post, Paul Krugman assumes a priori (he presents no theory) that greater government spending will lead to less unemployment, based on the current unemployment figures. So, I maintain my position.
That said, I never said that the empirical evidence “supports that stimulus is ineffective”. I believe that to be true, but the post explicitly says that I could easily fit that theory to the empirical evidence. It was not meant to be a proof of validity; I was making a point, which you apparently completely missed.