Monday, December 17, 2012

On the Nature of Forecasting Price Inflation and Other Economic Events

A troll has wandered over to my post of a Peter Schiff video interview with CNBC. The troll comments:
 Once more just the same old as far as forecast for a "dollar in freefall" or "prices going to skyrocket". When is this happening? How much "price inflation" as measured by the CPI of [sic] PCE index are we going to expect? Nothing but empty phrases which can mean anything and thus mean nothing. 
This type of comment comes across from time to time. It shows clearly commneters, who are completely unfamiliar with Ludwig Von Mises' Epistemological Problems of Economics and Friedrich Hayek's The Counter-Revolution of Science. Mises and Hayek made clear that the science of economics is different from the physical sciences. The nature of the science of economics is such that the exact forecasts that can be made in the physical sciences can not be made in the science of economics. Mises wrote in Epistemological Problems:
The specific experience with which economics and economic statistics are concerned always refers to the past. It is history, and as such does not provide knowledge about a regularity that will manifest itself also in the future.
But still we have trolls trying to bait Austrians to make exact forecasts. I hasten to add that because exact timing can not generally be known, it does not mean we can know nothing.

Consider a thought experiment, where a man stands dominoes on end, on a hardwood floor close enough to each other that if one falls a "domino effect" impacting other dominoes will occur.

Day after day, we see this man come into this room and stand on end more and more  dominoes. But soon we notice an open window with wind entering the room, some of the gusts appear strong enough to knock down a domino if it hits one just right. But day after day, the man comes in and stands even more dominoes on end, with the howling wind continuing. Now, we notice something else, young children are playing in the room, running around and oblivious to the dominoes. It is a big room and they are in another part of the room, but they are running, jumping, chasing each other, it is only a matter of time before they step on a domino that will start the others to fall.

Yet, day after day, the man comes in adding dominoes, some now closer to the open window, some now closer to the playing children. Is it not reasonable to warn that some, if not all, of those dominoes will come tumbling down? We can't say which day, because it is too complex to understand exactly when the children will play near enough or which day the gust will hit a domino just right with a enough force to start the chain reaction, but we see the dangers developing in front of our eyes.

To warn about such, if we are concerned about the dominoes falling, is an important warning. To dismiss the warning because the dominoes haven't fallen yet, or because we can't provide a certain time when they will fall or an exact number as to how many will fall, is absurd.

It's the same with the price inflation warning. The Federal Reserve is pumping more and more money into the system, this on its own would cause price inflation, if it wasn't for other factors such as the demand to hold cash and increases in productivity, which can on a short-term basis counter price-inflation trends.

But, we see Fed chairman Bernanke printing and printing, this is like the children playing in the domino room or the wind gusting through the open window in the room. It's not good for the dominoes and the Bernanke printing is not good for price inflation. At some point that money printing becomes tsunami like and overcomes the counter trends and the price inflation starts in earnest. Exact timing is difficult, if not impossible, to predict with accuracy, but we can see the problem ahead. To point out the price inflation storm that is developing is heroic, especially given the damage price inflation will do to those unprepared. To attack such a warning, because exact dates and levels of the coming price inflation are not given, is an attack coming from a clueless person, almost child-like in thinking, who doesn't understand how complex the world is and how it works. One, who is clearly not recognizing an important warning for a very dangerous developing price inflation storm. Ignore these trolls, prepare for the price-inflation that is coming.


  1. Brilliant analogy. well done!

  2. Logic says that producing more money would cause the price (value) of each unit to decline. That's because of the laws of supply and demand. Yet as Charles Hughes Smith said at Of Two Minds, we don;t really know the demand for US dollars.
    world trade is much greater after the fall of communism. There is more international specialization and international transactions in most production. Who knows how much additional international demand for dollars there is?

    Second is the experience of Japan. They have been easing for twenty years and interest rates/inflation keep going down.

    1. Each of those dollars is produced at a point, for benefit of the creator of that dollar. The extension that in the aggregate nominal prices must rise does not necessarily correspond, and this is where Walrasians and Friedmanites get into trouble. Rothbard explained well the idea that the movement of those created dollars into the economy is not a smooth process, but a disruption of the relative values of goods predicated on the preferences of the owner of those dollars.
      Japan may be a deviation from a Monetarist understanding of the ERE, but fully consistent with Austrian monetary theory.

  3. I would only note that ignorance doth not a troll make, and that government education a lot of ignorance doth make.

  4. The loathsome, vile, stupid and utterly dishonest MMTer Mike Norman accused anonymous Austrian commenters of predicting hyper-inflation.

    I pointed out to Norman that I'm basically the only Austrian who ever comments on his stupid blog and I've never predicted imminent hyper-inflation. I also linked to Bob Murphy's latest evisceration of MMT. Norman deleted the entire post and all comments within an hour.

    No anti-Austrian ever understands the problems of economic calculation and price distortions and no anti-libertarian understands the NAP. Ever.

    1. Your posts don't go unnoticed though, and you do a great job helping people understand that distinction. And your last paragraph is sad but too true.

  5. Wow, I really love that analogy. Thank you for that. It makes it so understandable. I hope you don't mind if it's "borrowed".

  6. Yes, pointing out that eventually WILL happen, ceteris paribus, drives them nuts. It's 1+1=2.

    You cannot continue to print money at a rate higher than the innovation level and expect no inflation. The fact that these guys will be shocked- crazy shocked- when inflation takes off, and interest rates rise just shows how warped their models are. A strong recovery (due to another manipulated boom) will drive the banks to move (currently sterile) assets from the Federal Reserve to the "real" economy, and expand the boom. It will be short lived, and once the high inflation creeps through it will crash again.


  7. As other commenters have said: Great analogy.

  8. Bob,

    In the passage by Mises where you quote comes from Mises states:

    "Historical experience does not provide facts that could render in the construction of a theoretical science services that could be compared to those which laboratory experiments and observation render to physics."

    This suggests that to a degree, as far as physics, we can learn from history. Hayek following Karl Popper came to reject this idea. That is, Mises continued to accept that so called inductive logic at least for the physical sciences was appropriate. But Hayek actually rejected inductive logic for even the physical sciences. He revised his opinion over time.

    The problem that Hayek sees with social sciences is *only* in that they are far more complex. For every possible falsifying bit of evidence, it's too easy to point out another factor and use this as an excuse as to why the theory didn't appear to hold – just as in your own example here.

    While Mises seems to suggest tautologies are okay in science, Hayek rejects this following Popper.

    Specifically Mises points out math and logic are surly sciences, right? Economics is a bit like that … but this misses Popper's point which was really stated by Einstein before Popper said anything about it:

    "As far as the propositions of mathematics refer to reality they are not certain, and so far as they are certain, they do not refer to reality." —Albert Einstein

    This point seems to be lost on Mises. But it's basically a summary of Popper's epistemology, which Hayek adopted.

    The best place to read about this is in Hayek's "Degrees of Explanation", where he specifically argues economic theories are *uncertain* and have an empirical content. Especially see section VII of that essay. Hayek would say we *can* and *do* make predictions, but it's simply far, far harder to keep a scorecard because of the complexity involved.

    So, of course, Hayek would readily agree with you as far as making predictions. But Rothbard and Mises seem to argue economic theory simply cannot, even in principle, be subjected to tests. Hayek would argue in principle we can *test*, only it's just extraordinarily difficult.

    1. Popper was right, in some ways. The actions of 7,000 humans, let alone 7B humans, cannot be known beforehand. Therefore models that show what happened before, and why one policy or another failed, or succeeded in examination of future policy is valuable.

      Good show!

  9. Well thank you for at least providing a post on why you do not want to be more specific, I appreciate this (without irony). I don't think comments such as "troll", "almost child-like thinking" or "clueless person" are warranted, but ok, I appreciate you taking the time.

    You can of course, as you know, come up with analogies for anything. Krugman and many others you (and I) might abhor also use them. They do not replace theory (and in my opinion) forecasts.

    I find it troubling that many Austrian economists return time and time again to the (undoubtedly correct) forecast of the housing bubble which many (including Mr. Schiff) made. But at the same time these very same people refuse to provide any substance to "the developing price inflation storm". While this is again a nice metaphor it leaves me wanting to know more. You criticized Krugman recently, appropriately I must say, on declining that Mr. Schiff forecasted a housing price bubble and collapse.

    To continue with your own analogy. How close the dominoes are to the window and the playing children clearly matters. The closer the higher the probability for them all falling; you seem uninclined to even wanting to measure this distance. Shouldn't this be the first task for everybody? Ben Bernanke would argue that they are a safe distance away and he is very much aware of this problem. He might claim he has tools in place, maybe already dangling from the roof, which might immediately be able to put safeguards (interest on reserves, reduction of balance sheet, forward guidance) between different sections of dominoes to prevent more from falling at the first hint of a child or wind knocking one over. Thus, the room even with lots of dominoes remains perfectly safe.

    1. Wonderfully thoughtful reply. I've pegged you as a troll (assuming you are the contrarian on similary posts) but apparently not.

      RW is positing that by setting up 100 dominoes, the ripple will exist, but be too small to measure. When there are 1,000 dominoes, it will be harder to keep them upright, and easier to measure. With 10,000 dominoes it will be harder still to do all the latter, and at the same time any toppling will have outsized effects, even if the first domino is toppled near the end of the chain.

    2. Now, imagine that the domino layer has constantly stopped those dominoes from falling, especially if one of them will knock down 2 or 3 dominoes (you've seen the layout- one domino can knock down a chain starting 3-5-7 more chain effects.) and suddenly he cannot reach out fast enough to stop those three chains. He has seen the first layer fall, even if knocked down halfway, and stopped it and reset the dominoes. But one day he comes in and the kids, or the wind, has started a chain reaction where not one but several different chains of dominoes are falling, and he tries to stop ALL of them from falling.

      That is the Federal Reserve in a nutshell.

  10. Second part due to character limit cut off before:
    You state clearly that a "price inflation storm is developing". In your analogy you and many other Austrian economists have been warning about dominoes falling for years. I would like to know what a "price inflation storm" looks like. Is it 5%, 10%, 100%, 1000% "price inflation"? And for me personally, it matters quite a bit to get at least a vague sense whether this (whatever it may mean in numbers above) is unlikely (say 10% probability), somewhat likely (say 33%) or very likely (say 66%) next year. My own forecast is 0.001% probability for 10% or more "price inflation" as measured by the CPI or PCE index (sorry for the typo in my quote) next year.

    That would also enhance the opportunity for your readers to take actions appropriate to your warning. Otherwise what's the use? Maybe the Fed is completely aware of what you call counter-factors and does appropriate policy to fulfill its mandate? After all "price inflation" has been remarkably close to the target of 2% for the PCE index. Private sector forecasts don't seem to indicate anything different for at least the next decade. The timing forecast also does not need to be exact. If you think there is a significant risk for next year that would be helpful if you could simply define what significant means in this context. You did not shy away from giving Geithner 75 to 1 odds of becoming Fed chair in 2014. So it seems you are not totally opposed to working with probabilities or making forecast.

    For me personally, a school of Economics (or social science more generally) loses much of its appeal when it can never be falsified. Mainstream Macro had lots of egg on its face with underestimating financial frailties in much of its forecasts. But your forecasts of "developing price inflation storms" should have an expiration date and a clear definition of storm or strong or whatever qualifier you use. If you want to be taken more seriously by many more beyond the core of "true believers" that to me is the test. Otherwise its the "boy who cries wolf" or the "broken clock which is right twice a day" or whichever metaphor you prefer. Your call. It might be helpful to know at which point you would question the validity of your theory. What if "price inflation" stays around 2% for another year, 5 years, 10 years, 20 years? That clearly matters for most investors time horizon. Many who predicted "significant price inflation" have missed a very strong bull market in bonds.

    1. Google "Thymology" and see why specific date forecasts are anathema to Austirans.

      The markets can remain irrational far longer than you can remain solvent.

      2008 should have been the end. If not for extraordinary and unprecedented intervention by the central banks, the entire system would have failed.

      God willing, the next failure will be too extreme, and they will go away. It will be rough, and painful.

      But, the world will be MUCH better once a sound fiscal system, absent central banks and oligarchs and feudal lords...I mean congress/presidents... Are gone.

      Got gold?

    2. Dale, thank you for your kind response.

      I agree with your quote an market irrationality, ironically it is by Keynes.

      Even though this takes the discussion away a bit from forecasting inflation, I don't accept the need to massively purge the system. It didn't want 2008 to be "the end". I don't want to witness the complete meltdown of the US financial system and economy. I don't believe that the Fed's actions have made a bigger crisis more likely. I do believe the Fed has the tools to combat inflation IF (a big IF in my opinion) it were to rise. Central banking (and academia) has spend the last decades talking about nothing but fighting inflation through central banking tools.

      Even though the analogy might be getting a bit tired, all central bankers have watched rooms with dominoes and kids and windows their entire life. They have thought about the timing of potential falls and the risks. The dominoes right now also fulfill a very useful role in limiting further damage (most on this blog probably disagree here) to the US economy by supporting growth. The blog's author is referring to the slowly recovering housing market. That's good! Not a sign of danger to come (if monitored appropriately).

      Currently we have low and stable "price inflation" expectations. We have high unemployment which research shows is NOT structural (see for example Ed Lazaer's recent work). This argues according to most mainstream economics models that an increase in "price inflation" is not likely at all. We have moved further away from the kids and windows rather than getting closer.

      You might disagree with this assessment. So give me some concrete evidence and forecasts about where I am wrong and how an impartial referee might settle the debate. I believe theory and forecasts can accomplish this. But only together. Austrian theory is well known, but without forecasts it can't be useful for policy making, in my opinion.

      As a final aside (yes, this is a bit of petty gotcha on my part): The author of this blog did predict double-digit for 2011 about two years ago:
      In the discussion about John Williams, Mr. Wenzel states, "I fully expect double digit inflation by the end of 2011, but we have a ways to go before hyper-inflation, which I would classify as inflation over 30%."

      These are the kinds of forecasts which are helpful to me. It didn't happen. It quantifies the terms hyperinflation as well. Even using the extremely flawed and misleading measure of "price inflation" employed by Mr. Williams, it didn't happen.

    3. The difference in philosophy displayed between your two arguments (thoughtful and well-written, both)represents an unbridgeable difference in thinking, it seems.

      Dale, I think, feels that the resources of a given society can't and shouldn't be top-down planned, by any group no matter how wise or academic, without violating the NAP (sad) and ultimately succumbing to the problem of economic calculation as formulated by Mises (inevitable though hard to predict when or how).

      Anonymous, I think, seems to accept the benevolence of wise and academic-types who can and should plan the distribution of a societies resources, especially in order to hedge against the instability or Darwinian-neglect of a free-market.

      I can rarely stand to read Keynesian/Austrian economic arguments because it tends to devolve from dialogue to idealogue. If you ask me, the argument necessarily has to go there as they both reflect contrasting philosophies for arranging society; but the arguments stated here are clear and lack the usual vituperation. So cheers!

      The issue, as I see it, between the two philosophies is the expectation of either party. Austrians tend to believe that people left to their own to engage in cooperative transactions will self-regulate an economy and distribute resources in ways the benefit the most people. Keynesians tend to believe that free-markets will inevitably benefit some at the expense of others, and hence, necessary allocations of a societies resources against the wishes of some of it's inhabitants remains a necessary evil, even if it abounds with unintended consequences.

      I don't really have anything to add, just my ongoing observations in the debate about how a society should be planned: hierarchy or spontaneity.

      I would like to think that once people see that the Austrian economic argument stems from the NAP that those same people would realize that it is in their best interests in the same way their current understanding sees Keysianism in their current interests.

    4. In response to tJm:
      I agree that much of the disagreement comes down, ultimately, to philosophical disagreements which will be (almost?) impossible to bridge. In my thinking, which I outlined here above and in comments to other posts, there is the need for central banks issuing fiat money with the goal of stabilizing output. Many here disagree and apparently see an inherently much less stable outcome with policy intervention than in their absence.

      What I feel is useful when talking to each other (rather than yelling past each other) is to find common ground. How best to measure consumer price inflation for one. I feel strongly that the current measures provided by the BLS or the BEA are much superior to other alternatives. I have laid out my reasoning for this in other posts.

      I feel it is also important to realize that the Fed's goal is limiting increases in "price inflation" to 2% as measured by the overall PCE index (INCLUDING food and energy). Much variation in this would be bad (both higher but also LOWER "price inflation") for businesses and households, increasing uncertainty greatly. It seems that many here are concerned much more about "price inflation" compared to "price deflation". You may think this is the wrong goal and a discussion on this is very much appropriate. That's one of the reasons I read some of the threads here. But this needs to be more than claiming something as inherently "unfair" (for whom??) or ignoring the empirical backing for the stated goal, which is widely available.

      To me personally, the financial crisis forced me to rethink many of the (failed) results of mainstream macroeconomics. I decided that I had made the mistake of not listening enough to critics of these positions. But to be able to do so I need not just a look back at history (where you can always explain everything with special factors or whatever) but also a forecast which uses theory as an application for policy and economic behavior. I find this lack of forecasting particularly unappealing.

    5. You cannot forecast the outcomes of nondeterministic models. Don't play this loser's game.

      The risk models used in the mainstream ignore wealth transfers due to Cantillon effects, and they also ignore the increased risks due to a rigid top-down central planning that is prevented from adapting to new information.

      Use the above two facts to aid your investing, and don't fool around with making predictions.

  11. Let's put it this way. money stock is as to potential energy what money velocity is to kinetic energy.
    Until demand picks up, "kinetic energy" will remain low.
    What Peter Schiff is saying is that the federal government's policy is that of a car climbing higher and hire up a steep and windy mountain road at faster and faster speeds. The car is gathering potential energy. Eventually the car will either reach the summit or fall off the side of the road and down the mountain cliff. Neither he nor Ben Bernanke nor anybody else knows when that point will come. However, the longer this money printing continues, the more gruesome the crash will be.

    1. Great analogy. I think our so called "troll" is just someone smart enough to look "behind the curtain" and will wake up to the beautiful and rational nature of Praxeology.

    2. It does seem that way, Dale.
      He's really hung up on replicating a physical science though,... and then suggesting a dangling thing mixed with children playing as if nothing could go wrong? That was funny.

      And this was an excellent blog post.

      - IndividualAudienceMember

  12. There is a book called "Future Babble" worth reading. The main premise is that we, as a people are addicted to someone telling us what will happen in the future so sooth-saying has become a lucrative business. But the more certain someone is about their prediction, the less likely it is to happen (roughly speaking). Most things really can't be predicted with the degree of certainty and accuracy that economists and social scientists claim and that most mainstream predictors gain "credibility" simply by stating their predictions loud enough, often enough, by sounding smart, and by altering their predictions after the fact. There is a chapter where the author deals with Peter Schiff's housing bubble prediction. His stance is that Peter had been saying the same thing for years and when "coincidentally" it finally came true, everyone thought he was a genius. Of course this fits into his neat little theory as long as he ignores the fact that (as far as I know) Schiff didn't put a time frame on his prediction.
    There's also a glowing chapter on George Soros' predictions (and investments), which I think is an entirely separate issue. When you have power to control or guide the machinations of states, it is easier to predict short term outcomes.

  13. There are many events taking place right now that are allowing the dislocations to continue, expand, and extend in ways that have never happened before. Never before has a government essentially nationalized the entire mortgage complex (it would be a stretch to call it a 'market' at this point...). Never before has a government essentially nationalized much of the student loan and educational complex.

    What happens when the government unwinds these positions? Can anyone even begin to conceive of how such a thing can be accomplished without huge dislocations, dislocations that get bigger and bigger every day and the fallout from which will be worse and worse? What other distortions can the government engage in to keep the house of cards standing a little while longer?

    You know what they say about inventing something that is supposedly foolproof: they'll just go and invent a greater fool...

  14. It would seem that Heisenbergs uncertainty principle might apply loosely to the problem of economic obeservations and predictions. We can get economic specifics, but only for the past and not for the future. We can tell where an economy is generally headed, but we can only due so at the expense of specifics