“Inflation is always and everywhere a monetary phenomenon”

Almost, but not quite.

Back in the days when dinosaurs roamed the earth, and Cambridge economists kept guard at the Temple of Keynes, Milton Friedman’s focus on inflation as a monetary phenomenon was a revelation—and an excellent one. Next to Joan Robinson’s surreal claims that printing money was not responsible for the German hyperinflation, Friedman’s version of monetary economics provided a very healthy dose of sanity. And as central banks across the world learned from the mistakes of the 70s and brought inflation under control, it became clear that the monetary authority indeed had the power to contain the price level via control of the money supply.

But it’s important to know what this account leaves out: how, exactly, do prices adjust? And over what length of time does this happen?

The modern view, backed up by impressive (though not entirely conclusive) empirical evidence, is that most prices are “sticky”, meaning that price-setting is staggered over time. Of course, not all prices adjust in the same way: some are especially sticky, while a few change almost every day—gas prices being the most salient example for consumers.

What happens if, say, there’s turmoil in the Middle East and oil prices suddenly shoot up? We see a sudden uptick in many conventional measures of inflation (though not necessarily in the ones that are relevant to macroeconomic stabilization).

But does this have anything to do with monetary policy? Not really. What’s happening is a relative price change: a supply shock changes the equilibrium price of oil worldwide, and oil is suddenly worth more vis-à-vis other goods and services. Since the price of oil is far less sticky than the price of other goods and services in the economy, this relative adjustment manifests itself in nominal prices in a one-sided way: other prices stay more or less the same, while oil prices skyrocket to whatever level is necessary to clear the market. We see an increase in the overall price level.

And in the short run, this is an increase that the Fed is almost powerless to stop. Unless it sets policy to be overwhelmingly tight—pushing the fed funds rate through the roof, employment consequences be damned—it can’t induce prices other than oil to fall quickly enough to offset the increase from oil over a short timeframe. As a result, we see occasional sudden spikes in the CPI (though also sudden drops!).

This isn’t some failure of monetary policy: it’s the natural consequence of a world where some prices are sticky and others are not. Any attempt to undo it would force the Fed to make wild adjustments in monetary policy, leading to a far less stable economy for all of us.

Advertisements

29 Comments

Filed under macro

29 responses to ““Inflation is always and everywhere a monetary phenomenon”

  1. wh10

    Good points here.

    Per hyperinflation, here is an alternative perspective from the MMT camp, which I hope doesn’t bother you too much – http://pragcap.com/hyperinflation-its-more-than-just-a-monetary-phenomenon/comment-page-1#comment-48239

    Speaking of, curious to hear your response to the most recent posts in the MMT entry (from Scott F and JKH). Opinion changed at all or is still bunk?

  2. Nathan Tankus

    “Back in the days when dinosaurs roamed the earth, and Cambridge economists kept guard at the Temple of Keynes, Milton Friedman’s focus on inflation as a monetary phenomenon was a revelation—and an excellent one. Next to Joan Robinson’s surreal claims that printing money was not responsible for the German hyperinflation, Friedman’s version of monetary economics provided a very healthy dose of sanity” up until this point i’ve enjoyed your posts. this post, and especially the quoted section, is abysmal. first it isn’t joan robinson’s claim, it’s the 20’s keynes structural argument. he showed, i think convincingly, that hyperinflation, and the hyperinflation in germany, occurred because of balance of payments problems resulting from a chokingly huge amount of foreign denominated currency debt and not simple money printing. there are only 3 ways for a country with foreign currency debt to pay it off, earn foreign currency through exporting, borrow more or sell assets. keynes convincingly provided the math to show that exporting more goods would hit a stagnant market, force down export prices and may even bring in less foreign exchange then before (interestingly, his analysis was very similar to john stuart mill’s arguments from a century earlier). this left borrowing more money, obviously unsustainable, and selling assets. foreigners did take ownership over lots of companies, land and capital assets but paying off the reparations debt could hardly be accomplished by selling off the whole country. this left money printing. they attempted to pay off their debts by buying the foreign exchange with printed money, which of course forced down the exchange rate. add to this that once they defaulted on their debt their industrial sector was literally military occupied, so that they kept on paying workers even though they were no longer producing, and you have hyperinflation. none of this however was simple money printing. it was a balance of payments crisis caused by tremendous foreign denominated debt a shut down of production. to claim that hyperinflation doesn’t fit this story well or Milton Friedman’s theories explained the data better, tellingly without actually presenting keynes actual argument, is intellectually misinformed (i would say dishonest but i would like to give you the benefit of the doubt). you can disagree with Keynes argument. What you can’t (or at least i wish you couldn’t) do is insult one side of the debate without bothering to have the slightest clue of what the opposing argument was or is.

    • wh10

      Agreed. There is a lot of nuance and historical complexity associated with every case of hyperinflation, and as a result, while money printing appears to be the cause because it occurred, the true roots of the problem are found in some type of exogenous event (e.g., war, regime change) and/or ceding of monetary sovereignty (foreign denominated debt). The link I provided touches upon this.

      Nonetheless, the point about relative price changes and sticky prices is an interesting, valid observation.

      • Nathan Tankus

        yeah i agree. i wasn’t trying to say his ultimate point was somehow invalidated by keynes, just that the way he set up this post was highly misleading. Even if he disagrees with keynes analysis, he should give it legitimacy as an analysis, and not just “a surreal claim” of a “dinosaur”

  3. Once upon a time, the economics profession viewed the arguments of Keynes and Robinson about inflation as central to their understanding of monetary policy. They thought inflation resulted only from factors like “excess demand” (whatever that was supposed to mean*) and external imbalances, not from monetary policy directly. Central banks based their policy on this understanding, and used massive Keynesian econometric models to predict the impact of their decisions on the macroeconomy.

    The result was a massive increase in inflation rates across the developed world in the 70s.

    Then, economists shifted to an understanding of inflation that placed a far greater emphasis on monetary policy. A few central banks implemented Friedman’s actual suggestion of targeting money aggregates, though it turned out that this was not the best way to think about monetary policy. Instead, most adopted the “Taylor principle” that central banks should control the money supply such that nominal interest rates increases more than one-for-one with inflation.

    The result, as we now know, was an unprecedented era of stable and low inflation. So even if the Keynes/Robinson view of inflation was theoretically plausible, it has been overwhelmingly refuted (at least as a practical way of thinking about monetary policy) by the experience of dozens of countries. In my view, you need some extraordinarily good arguments to overcome this empirical experience.

    *In some sense, “excess demand” is still the channel through which economists think monetary policy impacts short-term inflation: loose monetary policy lowers short-term real interest rates, pushing up demand and causing firms with some degree of market power to raise their prices. They just think that in ordinary times, it’s better to look at inflation and interest rates directly, rather than placing all the emphasis coming up with a model for “excess demand”. (which, like all econometric models, can never be entirely accurate)

    • It is unfair to say that economists don’t know what cost push inflation is.

      The inflation of the 70s was not due to a lack of theory. There was no political will from either the Johnson or the Nixon adminstrations to raise unemployment in order to deal with inflation which is the ugly trade off of cost push inflation.

  4. wh10

    I think a distinction needs to be made between an understanding of inflation vs hyperinflation. Two different animals. There is where the disagreement was above.

    • Has there ever been a case of hyperinflation where the government wasn’t trying to print money to keep itself solvent?

      • Nathan Tankus

        this is besides the point. it’s not money printing causing hyperinflation. it’s a balance of payments crisis causing it. calling hyperinflation a money printing problem is akin to calling a bullet wound a bleeding problem. yes a body will hemorrhage blood when it’s been shot but that’s a response to the bullet, not the root cause in itself. lets say that germany hadn’t printed any money once it hit balance of payments problems, would it not have had gigantic unemployment and a currency collapse?

      • wh10

        The point we’re trying to make is that, in these instances of hyperinflation, the money printing didn’t occur in a vacuum. Rather, it was preceded by some sort of extreme exogenous event, such as rampant political corruption, war, collapse of productive capacity, and/or the nation which experienced hyperinflation was not monetarily sovereign (e.g., it operated as a fixed exchange regime, its debt was demoninated in a foreign currency, etc.).

        Thus, the conclusion that it was simply the money printing which resulted in hyperinflation completely misses the bigger picture. Yes, one could conceive of a scenario in which a country randomly decides to print unlimited amounts of money, resulting in hyperinflation, but that wasn’t the case in Germany, and I’m not sure that’s ever been the case.

        So, the point is that inflation, defined as a continuous rise in prices, needs to be discussed in a different light than hyperinflation, which is characterized by the public’s rejection of the domestic currency. These are two different animals.

  5. lets say that germany hadn’t printed any money once it hit balance of payments problems, would it not have had gigantic unemployment and a currency collapse?

    It might have had a partial currency collapse—but not nearly to the same degree as the one it experienced. And no, there would not have been any hyperinflation.

    If the fiscal authority in Germany was separate from the monetary one, there would have been no impetus to print money to pay the government’s debts—that would have been the fiscal authority’s problem, not the central bank’s. It would have been a severe problem, to be sure, but without a conscious decision to print money (rather than, say, explicitly raise taxes) to cover external debts, the money supply would have remained relatively constant and there would not have been hyperinflation.

  6. Thus, the conclusion that it was simply the money printing which resulted in hyperinflation completely misses the bigger picture. Yes, one could conceive of a scenario in which a country randomly decides to print unlimited amounts of money, resulting in hyperinflation, but that wasn’t the case in Germany, and I’m not sure that’s ever been the case.

    Sure. Countries only engage in out-of-control money printing when they are in fiscal trouble. I’m not claiming otherwise. I am claiming that the decision to address fiscal problems by printing lots of money to fund the government was responsible for hyperinflation. Many other countries have suffered under extraordinary debt burdens, external imbalances, and so on, yet managed to avoid hyperinflation. The reason is that they have kept the central bank (mostly) separate from the fiscal pressures facing the government, and it has been free to conduct monetary policy in a way consistent with more reasonable levels of inflation.

    • wh10

      I certainly agree that running the printing presses exacerbates the problem, no doubt. But the process by which the country begins to lose faith in the currency starts before excessive money creation. And even in this post, you concede that monetary policy is ineffective in the case of more conventional cost-push inflation. So how would monetary policy have otherwise salvaged a country which eventually suffered hyperinflation? Could you please provide some examples of countries which have avoided rejection of the currency in scenarios comparable to Weimar, Zimbabwe, etc as a result of shrewd monetary policy?

      • I could cite every instance in which a country has experienced extreme fiscal pressure or default but has avoided hyperinflation. Pakistan defaulted on its sovereign debt in 1998, for instance, but didn’t experience extreme inflation or wholesale rejection of its currency. The same year, Russia experienced an enormous default, which was associated with very high levels of inflation (partly due to the collapse in the ruble, partly due to otherwise irresponsible monetary policy) but not hyperinflation (which, ironically, happened during the early 1990s, before the epic default). Indonesia experienced an epic financial and fiscal crisis that led to a collapse in the exchange value of the rupee and a brief bout of very high inflation (77% in 1998), but again not hyperinflation.

        It is very simple for a central bank to avoid hyperinflation if it exercises a modicum of discipline and is not forced to raise money to fund the government. Even if the “country begins to lose faith in the currency”, there won’t be hyperinflation (or even a high rate of inflation) if the monetary authority acts responsibly. As long as there is some demand for the currency (which there always is, even in the early days of a hyperinflation), the central bank can simply claw back the money supply until supply and demand for the currency match at current prices.

      • wh10

        Matt, I knew you would come back to school me :). Point is certainly well taken.

        Would you be able to expound upon the mechanism by which “the central bank can simply claw back the money supply until supply and demand for the currency match at current prices.” Certainly the central bank can alter interest rates to impact economic growth, but only taxes can literally reduce money possessed by the private sector.

        Lastly, if the actual quote is “The recognition that substantial inflation is always and everywhere a monetary phenomenon is only the beginning of an understanding of the cause and cure of inflation,” I am more at peace, though I recognize money printing is a key ingredient for hyperinflation.

      • You guys are missing the one big ingredient in the hyperinflation story, which is that you also need a massive collapse of your productive capacity. Comparing pakistan to russia, germany or zimbabwe is apples and organges becuase Pakistan did not have its industrial base occupied, nor did it exile its main producers nor did it lose 50% of its GDP inthe course of 2 years. The collapse of productivity is the mechanism that allows prices to rise as dramatically as they do to kick off a hyperinflation. i.e. the aggregate supply cuve shifts left so that you are already on the vertical part of the curve.

        Also, no amount of central bank independence can overcome the kind of fiscal trauma experienced by germany in the 20s.

  7. Scott Sumner

    Matt, I agree with the post, but I’d like to correct a couple historical points:

    1. Friedman said “Persistent inflation is always and everywhere a monetary phenomenon.” He certainly understood the impact of supply shock. So the first line of your post shouldn’t be “Almost.” He would agree with you.

    2. Don’t lump Keynes and Robinson together on hyperinflation. Keynes hated a pure fiat money; called it the worst possible monetary system, even worse that a rigid gold standard. He wrote a book in 1923 that discussed the German hyperinflation. As far as I know Keynes never applied his notion of monetary policy ineffectiveness to fiat money. He did not deny that a fiat currency could be debased whenever the government wanted.

    • Scott,

      Thanks for the comment. Regarding Friedman, I don’t doubt that he would agree with the sentiment here in many ways, and that his main point was that “persistent inflation” was always and everywhere a monetary phenomenon. But I’m a little surprised to hear that this was actually part of the famous quotation–are there any references for this? A Google Books search for the phrase turns up countless citations, but none of them seem to include “persistent” at the front of the quote.

      • David Laidler made the “persistent inflation” distinction in “The Illusion of Wage and Price Control”.

        I don’t know if that is a sufficient proxy for Milton Friedman’s view.

  8. Jim Glass

    “The recognition that substantial inflation is always and everywhere a monetary phenomenon is only the beginning of an understanding of the cause and cure of inflation.”
    –“Money Mischief, Episodes in Monetary History”, by Milton Friedman, p 193, as per Google Books.

    I looked there because I read it but don’t have my copy with me. My unsupported memory is that in it Friedman cited the popular quote as attributed to him, said he might have said it, then expanded on the subject saying prices can change modestly due to changes in demand for money and other causes but larger scale inflation is always monetary.

    In any event, however faulty my memory may be, that’s the quote from the book as per Google.

  9. Scott Sumner

    Jim Glass, Thanks. I don’t have the citation, but am told that Friedman was misunderstood. My hunch is he did once say what Matt asserted, but when people pointed to temporary non-monetary factors, later said something like “I meant persistent.” I know he clarified this at some point.

    • Jim Glass

      “My hunch is he did once say what Matt asserted”

      He probably said it lots of times — especially after it became a popular saying and people asked him about it. He lived a long life!

      I’m pretty sure I’ve seen him say it casually in video clips of conversations and such at YouTube. But when writing for publication one is more careful than when talking offhand.

      Again, my memory of the book is he brought up the unqualified quote as being attributed to him, said he didn’t remember saying it but might have, then adopted it as being correct with the necessary qualification as explained. Pretty much your scenario. In other situations I don’t doubt he used “persistant” or some other qualifier as appropriate for the audience — or not, if he was just being casual.

  10. Scott Sumner

    This has more information.

    http://economistsview.typepad.com/economistsview/2007/04/money_and_infla.html

    He may not have used the word “persistent”, rather in context he was discussing ongoing persistent inflation episodes, as opposed to one-time price level changes.

  11. Would you be able to expound upon the mechanism by which “the central bank can simply claw back the money supply until supply and demand for the currency match at current prices.” Certainly the central bank can alter interest rates to impact economic growth, but only taxes can literally reduce money possessed by the private sector.

    This can be done with either taxes or open-market operations. (In fact, in the US, there is an partial operational divide between the Treasury and the Fed, and changes in the supply of base money take place mostly through open-market operations rather than taxes: reserves leave the private sector and go into the Treasury’s general account when taxes are paid, but the Treasury’s general account is never kept at a very high level and the work of adjustment is done by the Fed. The Fed is not even allowed to buy Treasury debt directly; it has to go through primary dealers first. The lines have blurred in the last few years, admittedly, with the creation of the “Supplementary Financing Program”.)

    Now back to Germany and hyperinflation. If the central bank has enough in assets to back its currency liabilities—in other words, if it’s not insolvent—it is easy to prevent hyperinflation. Even if there is a loss in confidence and the amount of currency demanded at any particular interest rate declines—let’s say by a factor of 2—the central bank can just sell off half of its assets and restore equilibrium at the existing price level and interest rate.

    If the central bank is insolvent (perhaps because it’s invested in domestic government debt that will not be repaid), then life is more complicated, because in theory a sufficiently large loss in confidence would be self-fulfilling, as the central bank couldn’t sell enough assets to soak up the excess demand. I don’t think this is actually what happened—people need real money balances for transactions, and if the central bank has been responsible about keeping the total supply of base money in line with demand at stable interest rates, they’re unlikely to avoid carrying cash because they suddenly “lose confidence”. But it’s possible.

    As long as the central bank’s liabilities are a relatively small portion of overall government liabilities, however, even insolvency shouldn’t be too much of a concern: the fiscal authority can swap reserves in its account (obtained via taxes, as you suggest) with government bonds held by the central bank at par, allowing the bank to mop up its currency liabilities. This might be painful for a fiscal authority that’s already in bad shape, but if the currency liabilities aren’t very large it won’t be crippling relative to all the other fiscal obligations.

    Maybe this wasn’t possible in 1920s Germany (back then, technology and banking were primitive enough that paper currency was a pretty important asset relative to the rest of the economy), but it’s definitely true in a lot of modern countries. There are a lot of dollars in circulation outside the US, for instance, but even then the total quantity of currency is only $1 trillion, which is a *lot* less than the US’s GDP or total debt.

    • wh10

      Thanks for the reply.

      “This can be done with either taxes or open-market operations. (In fact, in the US, there is an partial operational divide between the Treasury and the Fed, and changes in the supply of base money take place mostly through open-market operations rather than taxes: reserves leave the private sector and go into the Treasury’s general account when taxes are paid, but the Treasury’s general account is never kept at a very high level and the work of adjustment is done by the Fed. The Fed is not even allowed to buy Treasury debt directly; it has to go through primary dealers first. The lines have blurred in the last few years, admittedly, with the creation of the “Supplementary Financing Program”.)”

      In the case of OMO, reserves leave the banking system, but that does not destroy deposits, currency, or any other net financial asset held by the public. So OMO cannot reduce the money supply; OMO just swaps reserves for treasuries- that is the accounting. Only taxes can directly reduce aggregate demand (in which case reserves AND deposits, for example, are decreased). OMO might alter the growth of money supply, but it cannot destroy it.

      As for the rest, I will need time to digest. Thanks for the discussion.

  12. TC

    This is off topic, but still related to your post. I’d argue that oil prices should never be a concern of monetary policy because the demand for oil is so price inelastic.

    If we are looking for people to begin to favor money over oil and gasoline, relatively large changes in price do not seem to have much impact on the demand for oil and gas. Why should we assume that changes in interest rates will have much impact on the demand for oil and gas?

    http://traderscrucible.com/2011/03/08/the-real-reason-to-ignore-energy-inflation-when-core-inflation-is-low/

    Then, even moderate changes in economic activity don’t seem to have much impact on the quantity of oil and gas demanded. The amount of economy-wide demand destruction would have to be large to have much impact on gasoline and oil prices.

  13. Chris

    “this relative adjustment manifests itself in nominal prices in a one-sided way: other prices stay more or less the same, while oil prices skyrocket to whatever level is necessary to clear the market.”

    I am curious how come you assume these “other prices” can’t fall in money terms. Surely, assuming stable demand/supply fundamentals of the medium of exchange, a spike in the oil price would lead to a concomitant decline of other prices, keeping the general price level unchanged.

    If these “other prices stay more or less the same, while oil prices skyrocket to whatever level is necessary to clear the market,” the markets for many “other goods” will be unlikely to clear. At “stable prices” there would be an oversupply of “other goods.”

  14. Reblogged this on Utopia – you are standing in it! and commented:
    Joan Robinson thought German hyperinflation was not caused by monetary policy!!

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s