Good deflation/bad deflation, good inflation/bad inflation

I recently attended an economic conference with mostly conservative-leaning economists.  Someone had a paper that mentioned how certain types of deflation can actually be good, as when rapid productivity growth helped reduce prices in the late 1800s.

I agree with this, and mentioned that I rarely hear conservatives talk about “good inflation.”  Well I might as well have thrown a skunk into the middle of the room.  Let’s just say that the idea of “good inflation” didn’t go over too well.

And isn’t that the problem?  Isn’t that why we are where we are?  We have all sorts of models that are basically symmetrical.  You might argue that a stable price level is ideal, and that any inflation or deflation is bad.  But if you argue that some deflation is bad and some is good, then you implicitly have a model that distinguishes between demand and supply shocks.  So supply or productivity-driven deflation is good.  Of course those models imply that inflation caused by a fall in aggregate supply is also good.  The models are completely symmetrical.   This shouldn’t even be controversial.

So why don’t conservatives look at things that way?  And why do you rarely hear liberals talk about “good deflation.”  Maybe it’s just mood affiliation.  Or maybe people just don’t have the right model in their heads.  (I.e. the model I have in my head.)  Some people do understand that the argument is symmetrical.  (I seem to recall both David Beckworth and George Selgin making similar observations.)  But it seems to me that they (and a few others) are the exception.  And maybe that’s why the Fed is having so much trouble creating “good inflation.”

PS.  Notice that NGDP targeting automatically creates deflation when deflation is appropriate and inflation when inflation is appropriate.


Tags:

 
 
 

72 Responses to “Good deflation/bad deflation, good inflation/bad inflation”

  1. Gravatar of Gregor Bush Gregor Bush
    14. December 2011 at 08:47

    Scott you nailed it.

    You should have a macro version of your “never reason from a price movement” saying:

    “Never reason from an AGGREGATE PRICE LEVEL movement.”

  2. Gravatar of Neal Neal
    14. December 2011 at 08:49

    From a crude model of people’s politics, conservatives represent the wealthy, saving class, so they object to inflation on self-interested grounds. Progressives represent the poor, consuming class, so they object to deflation on self-interested grounds. People will rationalize what is in their group’s best interest.

  3. Gravatar of Silas Barta Silas Barta
    14. December 2011 at 09:11

    Arrrgh. You’re confusing so many different issues here. Let me see if I can untangle the equivocations.

    1) There are two kinds of inflation (I’m using the term to include deflation here, which is negative inflation): money-driven inflation (MDI) and technology driven inflation (TDI). Negative TDI is good from the standpoint of consumer welfare. The central bank should prevent large positive MDI (see point 2 below for reasons). It should not care about TDI, as that’s orthogonal to its goals of making the right monetary policy adjustments. (For similar reasons, it should prefer core over total inflation as a measure.)

    It definitely should not count negative TDI against positive MDI so as to come up with low “inflation numbers”, which is what William Dudley did when he tried to claim that inflation (that the CB should care about) is low on the basis of better iPads. Bad, bad, bad.

    2) There are reasons that positive inflation (of either kind) is bad: because it takes from individuals’ savings, forcing them to move ahead purchases at the cost of their ability to maintain standard of living into the future. This inflation is always bad. However, certain things can *mitigate* the badness of inflation. I will list them below.

    a) If inflation coincides with a booming economy, and people are better able to find their comparative advantages and thus employ themselves producing valuable things, that can outweigh the badness of inflation.

    b) If after-tax interest rates fully compensate for the loss of purchasing power due to positive MDI, that can cancel out the harms of inflation. (Though obviously savers prefer the high interest rates to coincide with zero inflation.)

    c) If the inflation is due to supply shocks (which counts as positive TDI, or perhaps its own class), then than is bad from the standpoint of consumer welfare. However, it is not the kind of inflation a CB should try to fight (i.e., it’s not MDI). And this inflation is “good” in the sense of, “it’s better for such real factors to be reflected in prices than for them NOT to be reflected in prices, resulting in non-price/non-market rationing”. But again, it would be better for the supply shocks not to happen.

    None of the above implies, in any way whatsoever, that inflation is good, just that it can coincide with things that are better than inflation is bad.

    There, did I untangle all the issues and resolve your confusion, Scott_Sumner? What of the above do you disagree with?

  4. Gravatar of Andy Harless Andy Harless
    14. December 2011 at 09:26

    Symmetry makes sense, but why would the ideal necessarily be zero? The received wisdom holds that the ideal long-run inflation rate is around 2% (or just slightly under). If that’s the case, then good deflation will be quite rare, although it could happen occasionally in response to unusually large supply shocks. If one believes, as I do, that the optimal long-run inflation rate is more like 4%, then good deflation is unlikely ever to happen in the real world, although it’s still a theoretical possibility.

    The conventional view among central bankers today, though, seems to be inclined toward inflation targeting at 2% or just below. In that case, any inflation rate significantly below 2% is bad, and so deflation is always bad (just as inflation above 2% is always bad).

    Thus there’s nothing inherently inconsistent in the fact that liberals seldom refer to “good deflation.” It’s just that it’s (depending on which version of the liberal view you take) either very rare or impossible, while “good inflation” is both possible and likely. Why conservatives seem to believe that there can be good deflation but not good inflation is more of a mystery. Perhaps they think that the ideal long-run inflation rate is minus 3%. I can’t think of any good arguments why that would be the case, though. The closest I could come would be Selgin’s arguments, but Selgin apparently already admits the possibility of good inflation.

  5. Gravatar of Silas Barta Silas Barta
    14. December 2011 at 09:48

    @Andy_Harless: Why conservatives seem to believe that there can be good deflation but not good inflation is more of a mystery.

    And I believe I resolved your confusion in the comment immediately preceding yours.

  6. Gravatar of Ryan Ryan
    14. December 2011 at 09:56

    Isn’t this just reasoning from an aggregate price change? I thought we weren’t supposed to do that.

  7. Gravatar of JimP JimP
    14. December 2011 at 10:02

    It just seems to me to be obvious.

    Bernanke does not like deflation – but he is and will deliver deflation because if he delivers inflation, aggressive enough monetary policy to actually make a difference, he will be impeached. The Republicans have made that real clear. Of course he would not now be convicted – and if I were Obama that would be a political fight I would welcome. But Obama would rather run on spending us into the ground.

    Its just a political war. Its not personal.That is what politics just is. And as Morgan keeps saying, once the Republicans get in they will deliver the monetary ease they are currently denying us – just as Germany will if the Euro crisis ever starts to actually harm Germany, which is has not yet done.

    Is this hypocritical? Of course – and what do you expect exactly?

  8. Gravatar of John Thacker John Thacker
    14. December 2011 at 10:03

    I agree with Silas Barta. The reason conservatives can speak of “good deflation” is that supply or technology caused deflation is the result of a good thing. A fall in aggregate supply may necessitate inflation (and yes, to be sure, it’s also as difficult as pulling teeth to get conservatives to admit this), but it’s certainly never “good.”

    The logic is– Raise in Aggregate Supply == good, Fall in Aggregate Supply == bad. So we have a good thing that causes us to need deflation, versus a bad thing that causes us to need inflation.

    None of this changes the argument in favor of NGDP targeting.

  9. Gravatar of John Thacker John Thacker
    14. December 2011 at 10:07

    @Ryan:

    Actually, it’s the opposite of reasoning from an aggregate price change. By pointing out that deflation or inflation can be the result of two opposite things, Scott is stressing that one should not reason from the price change.

  10. Gravatar of John John
    14. December 2011 at 10:14

    I agree with you Scott. At least as far as goods production has the power to move prices one way or another. If a disaster were to wipe out supplies and the supply and demand for money stayed relatively constant, prices should rise. However to Austrians, this goods induced fall in purchasing power is not inflation. Inflation means a change in the purchasing power of money based on supply and demand for money. Rising or falling prices are just one of many consequences of changes in the money relation.

  11. Gravatar of Becky Hargrove Becky Hargrove
    14. December 2011 at 10:43

    The real issue for some conservatives – especially as many financial instruments have run their course in the present – is what inflation represents: growth itself. Conservatives have little reason to complain about monetary growth when it is easy to see places where growth might happen without representing a threat to those who already have resources. But the fact that there are few ready-made places for growth in the present make me question the willingness for conservatives to just ‘open the spigot’ in the next Republican administration.

    For a long time, the conservative businessman has received top-down pressures from government that continue to impact the ability to make a profit. And if that were not enough, as unemployment continues as a significant issue, they are also starting to get pressures to allow business formation from the bottom: i.e. the request to loosen regulations and zoning rules that would allow a lowering of the bar for business in general. How would they be convinced that those new businesses would not make it harder on them?

    The only way to make progress on that front is for those in power to come to agreement on the most basic issue of all: government is not going to be able to take care of us in the future. It still wants to but it simply cannot. When everyone gets that, it will be possible as well to finally tell the truth: people are going to have to start taking care of themselves, which means monetary, not fiscal policy. The conservative needs to remember that when business is not allowed to grow, eventually government does not even have enough money to wage war or keep the misfits behind bars, because too few people are able to contribute to government at all when they do not have work.

  12. Gravatar of Benny Lava Benny Lava
    14. December 2011 at 10:48

    Scott,

    My theory is that conservatives don’t truly believe in bad deflation. They think deflation is good. So to convince people of this you need to do two things. First, demonize inflation. Second, rehabilitate deflation. You were a witness to some good old fashioned deflation rehabilitation. If we start looking for good deflation we just might find out that there was more good deflation than we thought. Maybe it was all good?

    I believe conservatives refer to this overall strategy as the wedge.

  13. Gravatar of marcus nunes marcus nunes
    14. December 2011 at 11:00

    This from Earl Thompson:
    “Perhaps the most common objection to a simple monetary expansion has come from the large number of amateur and semi-professional economists, who argue that an increase in the money supply would be “inflationary.” Of course it would be, but that is exactly what has been needed all along to restore the early 2008 demand for labor and employment levels.”

  14. Gravatar of Steve Horwitz Steve Horwitz
    14. December 2011 at 11:00

    Exactly right Scott.

    Changes in prices that come from changes in supply/productivity are good – that’s what we want prices to do. If goods get more scarce, say due to a natural disaster, prices should go up. Call that “good inflation” if you want.

    It’s the “aggregate demand” or monetary-driven changes that are a problem, hence both bad deflation and bad inflation. And where NGDP and the Austrian/free banking crowd agree is that under good monetary institutions/policy, the price level should move with those supply side changes. And that’s a good thing.

  15. Gravatar of Benjamin Cole Benjamin Cole
    14. December 2011 at 11:01

    The obsession with nominal systems is destructive. Inflation or deflation is much less important than real growth.

    For peevish and unhealthy reasons, some economists have begun genuflecting to price stability or gold, or zero inflation (though even measuring inflation is very tricky, and subjective).

    Give unto me real growth of five percent a year for five years, and I will eat inflation or deflation for breakfast. So will our nation—boomtimes help everyone.

    Remember: The purpose of microeconomic policy is prosperity. Winning. Keep your eye on the ball—not the box scores. Having low ERA’s doesn’t help if you lose the games. Baseball analogies, sorry.

  16. Gravatar of Marcelo Marcelo
    14. December 2011 at 11:43

    Scott,

    I think that this kind of attitude represents the battles that were waged in the 70s and 80s over dealing with inflation, and anyone that went through that now finds themselves unable to be for inflation. Additionally, my feeling is that people assume inflation is some sort of genie that once let out of the bottle is very difficult to reign in. I don’t think anyone that feels this way is going to be able to accept the NGDP targeting idea because they feel central banks are unable to reign in inflation once it ‘starts going’

    I agree that this is a silly way of looking at the world, but these the prejudices and mindsets that exist with many, particularly conservative, intellectuals and I think you need to try to reason with them understanding these pseudo-models about inflation one of the first battles that conservatives won in the 80s, and thus they are going to be very unwilling to change their mind on its evils.

  17. Gravatar of Jason Odegaard Jason Odegaard
    14. December 2011 at 13:07

    They have a knee-jerk reaction against inflation not because of an rational economic model. The reaction is a moral one. In most conservative economists, the position of a saver is the morally superior position, and any policy choice that harms the morally superior group is viewed as the wrong policy. This is a postulate, no proof needed.

    This is an impulse from morality, not from economic efficiency. And economics is really about efficient outcomes, not moral outcomes.

  18. Gravatar of JimP JimP
    14. December 2011 at 13:24

    http://uneasymoney.com/2011/12/14/deutsche-bank-gets-it-why-cant-mrs-merkel/

    One might gather from this that even deutsche bank thinks keeping nominal spending steady is a good idea. Maybe Merkel is a depositor there and might read this.

  19. Gravatar of Morgan Warstler Morgan Warstler
    14. December 2011 at 13:26

    “Well I might as well have thrown a skunk into the middle of the room.”

    Because you are a horribly shitty arguer, who would sell a hooker by talking about her intellectual prowess… since that is what matters to you.

    You selling “good inflation” by saying that in the AGGREGATE over the last 20 years, your approach would have lead to overall significantly less inflation.

    You sell it as THE way to hack the Democrats agenda and turn the fed into the GOP grass roots bitch.

    Both are TRUE, but you refuse to cater to wat the people REALLY WANT.

  20. Gravatar of Jason Odegaard Jason Odegaard
    14. December 2011 at 13:29

    PS – Scott, you are more of a utilitarian, which is why you can have a fully symmetrical model. I suspect you would generally view the efficient outcomes of markets as typically being the moral outcomes. And frankly, most people would also agree with the end product is the moral one! But when people hear a moral absolute being broken (example, harm to savers), all agreement is lost.

  21. Gravatar of Morgan Warstler Morgan Warstler
    14. December 2011 at 13:29

    “Well I might as well have thrown a skunk into the middle of the room.”

    Because you are a horribly shitty arguer, who would sell a hooker by talking about her intellectual prowess… since that is what matters to you.

    You selling “good inflation” by saying that in the AGGREGATE over the last 20 years, your approach would have lead to overall significantly less inflation.

    You sell it as THE way to hack the Democrats agenda and turn the fed into the GOP grass roots bitch.

    Both are TRUE, but you refuse to cater to what the people REALLY WANT.

  22. Gravatar of Steve Horwitz Steve Horwitz
    14. December 2011 at 14:09

    I find it interesting that everyone commenting seems to know what “conservative economists” think, yet being a libertarian economist, my comment was in full agreement with Scott. Maybe you need to get over your stereotypes.

  23. Gravatar of StatsGuy StatsGuy
    14. December 2011 at 14:29

    Kudos to you. (no humor)

    “Maybe it’s just mood affiliation. Or maybe people just don’t have the right model in their heads.”

    Or maybe they are simply hypocrites who are goal-seeking a preferred outcome, just like their mirror images on the other side of the debate?

  24. Gravatar of Donald Pretari Donald Pretari
    14. December 2011 at 14:32

    It’s funny. Because what you argue is so obviously correct to me, I feel that I must be missing something.

  25. Gravatar of John John
    14. December 2011 at 14:59

    Scott is technically wrong from an Austrian point of view. Goods induced changes in purchasing power such as productivity gains or supply destruction don’t qualify as inflation or deflation. Inflation and deflation refer to changes in the supply and/or demand for money. This may seem like mere semantics but it becomes important when you want to talk about stopping inflation or deflation. If you think that inflation or deflation only refer to price movements you’ll end up attacking the symptoms instead of the disease.

    In the case of inflation or deflation (defined in the above manner), the price movements are the economy’s natural adjustment mechanism to increased or decreased volumes of spending.

  26. Gravatar of George Selgin George Selgin
    14. December 2011 at 15:19

    “Inflation and deflation refer to changes in the supply and/or demand for money. This may seem like mere semantics but it becomes important when you want to talk about stopping inflation or deflation.”

    With all due respect, Jon, what you claim as a “technical” point is in fact “mere” semantics; indeed it is rather less than that, because according to your definition to “stop” inflation and deflation isn’t possible unless you can somehow stop, not only changes in the money stock (M2? M1? MZM?, B?), but changes in the “demand” for money!

    While it is true, as Scott observes, that we mustn’t regard any and all inflation and deflation (in the terms’ usual sense) as undesirable, appeal to Austrian definitions doesn’t take us any closer to recognition of this truth. Indeed, conventional Austrian thinking (I am thinking here of the Rothbardians), while insisting on using “inflation” to mean “monetary expansion,” also holds all of this sort of inflation to be “bad,” which is (in my view) to commit an error even more serious than that of regarding all upward price movements as bad.

  27. Gravatar of Lorenzo from Oz Lorenzo from Oz
    14. December 2011 at 15:42

    Silas: don’t your arguments assume, at some level, a static economy? Or, at least, discount expectations about the value of money having an effect on total transactions. If inflation increases transactions, surely that is “good” inflation by any measure?

    After all, the point about inflation eroding (monetary) savings has the equivalent with deflation in the eroding the value of (non-monetary) assets.

  28. Gravatar of Silas Barta Silas Barta
    14. December 2011 at 15:51

    @Lorenzo_from_Oz: Thanks for the reply.

    My arguments don’t assume a static economy at any interesting level.

    If inflation increases transactions, surely that is “good” inflation by any measure?

    No, that means the goodness of the transactions (probably) outweighed the badness of the inflation. It may be tricky to appreciate, but your claim is like saying that if you burn gas to travel to somewhere you want to go, then that’s “good losing gas”. But no, losing gas is always bad, just not necessarily as bad as what you got in exchange.

    After all, the point about inflation eroding (monetary) savings has the equivalent with deflation in the eroding the value of (non-monetary) assets.

    Nobody owns a “non-monetary asset” per se. Just like nobody travels in a non-car, or paints their house non-white. Rather, the own a specific asset, which isn’t money. (Or paint a specific color, which isn’t white, or use a specific mode of transportation, which isn’t a car.) And those assets may increase or decrease in value during inflation or deflation.

    So people don’t have the option to hold “non-monetary assets” as a hedge against inflation, only to hold a basket of assets which aren’t money, in the hope that they will beat inflation against the wishes of the CB.

  29. Gravatar of Silas Barta Silas Barta
    14. December 2011 at 15:53

    @Jason_Odegaard: That’s quite a high horse you’re riding, for someone so opposed to moralizing.

  30. Gravatar of Becky Hargrove Becky Hargrove
    14. December 2011 at 16:26

    Steve Horwitz,
    I was trying to find an explanation that was not denigrating anyone but looking for a rational business explanation to the best degree possible. Basically it is hard to expect anything to change in the marketplace if it has to come out of someone’s ‘hide’, for understandable reasons. As a libertarian, I believe that if people can take back the responsibility of a considerable amount of the service sector, thus relieving business of that burden, that conservatives would then have a real reason to reconsider allowing the marketplace as a whole to change.

  31. Gravatar of Lucas Lucas
    14. December 2011 at 17:56

    @Steve Horwitz,
    “If goods get more scarce, say due to a natural disaster, prices should go up. Call that “good inflation” if you want.”
    This can only be called “good inflation” on efficiency grounds since it’s obviously bad on welfare grounds (i.e. less real incomes)

    @Donald Pretari,
    “It’s funny. Because what you argue is so obviously correct to me, I feel that I must be missing something.”
    I feel the same, too. It seems that NGDP targeting and the AD/AS model have brainwashed us 😉

  32. Gravatar of Andy Harless Andy Harless
    14. December 2011 at 18:02

    @Silas

    There are reasons that positive inflation (of either kind) is bad: because it takes from individuals’ savings, forcing them to move ahead purchases at the cost of their ability to maintain standard of living into the future.

    In my opinion, this is 180 degrees wrong. Anticipated inflation only takes away from individuals’ savings to the extent that they use money as a saving vehicle. We should be discouraging people from using money as anything but a medium of exchange. Now it has to maintain its value to some extent in order to be useful as a medium of exchange, and indeed, it will necessarily be used to some extent as a saving vehicle, since one has to maintain positive balances in order to use it as a medium of exchange. But by allowing people to use money deliberately as a saving vehicle, we are allowing them to deviate from the optimal free market solution. Instead of saving in the form of some actual thing, which would aggregate across individuals, people can save (individually) in the form of money, which is a phantom and does not aggregate across individuals, so that individuals are able to save without increasing aggregate savings. This wouldn’t be a problem if prices and wages were perfectly flexible (and not under the attempted control of a central bank), specifically because people wouldn’t be able to avoid inflation: if everyone wanted to save, then prices today would go down, so that they would rise in the future. But we have prices and wages that are sticky, and we have a central bank that targets a constant inflation rate, thus making prices and wages even stickier. Under these circumstances, people who save in the form of money are taking advantage of a loophole in the system, and they should be forced to give up that loophole and move ahead their purchases.

  33. Gravatar of Jason Odegaard Jason Odegaard
    14. December 2011 at 18:11

    @Silas_Barta

    Haha, yes that may be true! But that’s what I see in some of the tenor of discussion about inflation – especially visible in arguments over whether the ECB should take any action to ameliorate the euro crisis. But thanks for keeping me in check. Yet, I’m not against moralizing. I am just pointing out that moral arguments won’t easily bend to a some rational economic model.

    Anyways, I quite liked Lorenzo’s argument about inflation/deflation, how inflation degrades monetary savings but deflation degrades non-monetary savings. But when it comes to a counter-point, I don’t know if we can use personal assets like cars or paint to compare non-monetary savings. It seems we are talking about different forms of savings.

    You get to it in your last sentence – people hold a mixture of assets that are non-monetary. And, as you mention, the value of individual asset classes rise and fall relative to each other whether there is deflation or inflation.

    The issue with deflation (while NGDP growth is low or negative) is that it incentivises economy-wide holding of the monetary asset – cash – rather than non-monetary assets. And everyone preferring to hold cash instead of non-monetary assets is not an efficient economy-wide trend. And that’s where inflation can occasionally help improve the economic-wide outcome.

  34. Gravatar of ssumner ssumner
    14. December 2011 at 18:24

    Thanks Gregor.

    Neal. 2 responses:

    1. Superneutrality of money
    2. I’m a fat cat capitalist, and I would benefit from monetary stimulus.

    Silas, You didn’t address my point at all. The models are symmetrical. If there is good deflation there must be good inflation. It’s almost a tautology. More precisely it’s a tautology that if it’s sometimes good for inflation to be below average, then it’s sometimes good for it to be above average. Then the only debate is over where the average should be. Which brings me to Andy – –

    Andy, You said;

    “If one believes, as I do, that the optimal long-run inflation rate is more like 4%, then good deflation is unlikely ever to happen in the real world, although it’s still a theoretical possibility.”

    I agree with your logic, but not your judgment. I think that 10% NGDP growth target would be fine for China. I see no arguments that it’s too low. But that would have led to “good deflation” roughly 50% of the time over the past 30 years. Indeed I think 8% would have been fine, with 2% trend deflation. (Had to say that because George Selgin is watching this post) 🙂

    Of course a conservative could make an analogous argument using a trend rate of deflation derived from Friedman’s optimum quantity of money paper.

    Ryan, That’s exactly my point. We should target NGDP, as inflation might be good or bad. We should never reason from inflation.

    JimP, They’ll demand Bernanke produce inflation the moment they get in office, just like under Reagan.

    John Thacker, No, consider a natural disaster. Conservatives are always the ones insisting that it’s a good thing to let prices adjust upward, so there’s no rationing.

    Thanks John.

    Becky, Inflation isn’t “growth” if it’s supply-side.

    Benny, Maybe, But I like to take people at their word.

    Marcus, Thompson was right.

    Thanks Steve.

    Ben, I think we can get more growth, but still want to target NGDP.

    Marcelo, That’s what this blog is for.

    Jason. That’s moral reasoning based on 100% faulty logic. The average rate of inflation (and hence welfare of savers) is exactly the same under inflation and NGDP targeting.

    JimP, Thanks, That’s a hopeful link.

    Morgan, It’s hard to argue at conferences, I make my good arguments here in the comment section.

    Statsguy, I prefer to think the best of people (except Rick Perry.)

    Thanks Donald.

    John, See George Selgin, that’s my response too.

  35. Gravatar of Silas Barta Silas Barta
    14. December 2011 at 19:41

    @Scott_Sumner: I was not trying to address your bottom line point. I was simply trying to resolve your supposed confusion over the senses in which inflation can be good, or bad, or avoidable, or CB-controllable. Are you going to continue to claim confusion on the issues you listed, or do you believe that my post resolved them?

  36. Gravatar of John Thacker John Thacker
    14. December 2011 at 19:49

    No, consider a natural disaster. Conservatives are always the ones insisting that it’s a good thing to let prices adjust upward, so there’s no rationing.

    Nope, Scott, I still believe that conservatives say that we “have to” let prices go up, rather than calling it “good price-gouging.” The phrase “necessary evil” applies to both things.

    There are some libertarians who insist on calling price-gouging, and payday lending, and other things actually “good,” but conservatives tend to adopt the moral “it’s an evil, but a necessary one, because without it things would be even worse, and really it’s only a symptom not the root cause” framework.

    In my opinion your response only strengthens my (and Silas Barta’s) point.

  37. Gravatar of Silas Barta Silas Barta
    14. December 2011 at 20:04

    Jason_Odegaard & Andy_Harless: I think you have both highlighted the fundamental weak point in AD/monetary equilibrium explanations. Once you accept that money must hold its value over time in order to be a medium of exchange, then on what basis can you claim that it’s “okay” for people to hold value that way for *some* lengths of time but is somehow inefficient to do so for longer lengths?

    If it’s somehow sending an important market signal — and not in any way introducing inefficiencies — for me to hold my money a day before finding something worthy of purchasing, then why does this not hold for two days? Three days? A week, month, year, decade?

    People would not value money unless it permitted them to indefinitely defer deciding what they will redeem it for. If people already know what they’ll want to buy in exchange, then there’s no reason to have money — just use barter and forward contracts.

    Frankly, I prefer the simpler explanation that all decisions to buy anything or not are sending important, relevant, informative market signals that allow the pricing mechanism to work, rather than saying that one particular instance of it causes all kinds of untold chaos that can only be averted by “the printing money”.

    PS Jason_Odegaard: Props on the underscore!

  38. Gravatar of Neal Neal
    14. December 2011 at 20:32

    Scott-

    1) Fair.
    2) Only because you’ve patented NGDP targeting!

  39. Gravatar of Benny Lava Benny Lava
    14. December 2011 at 21:09

    Jason Odegaard,

    This is an astute observation and one I have seen myself. Savers are morally superior to debtors. Inflation is theft and therefore morally wrong always and everywhere. This is an argument I have read countless times in countless places. This is why I believe conservatives believe that deflation is always good and inflation always bad. It isn’t necessarily based on what outcome produces that greatest GDP growth, increased wages, decreased unemployment, increased prosperity, or whatever metric you use to measure economic improvement. It is based on the PRINCIPLE.

  40. Gravatar of Benny Lava Benny Lava
    14. December 2011 at 21:11

    Steve Horwitz,

    Maybe libertarian and conservative are not always the same thing (Though they usually are)?

  41. Gravatar of Rahul Rahul
    14. December 2011 at 23:21

    Is there a defining Scott Sumner post outlining how NGDP targeting would work? I’m joining in the debate late and most mentions of NGDP assume away a lot of stuff.

    Any pointers?

  42. Gravatar of mbk mbk
    15. December 2011 at 04:29

    It seems strange that no one has mentioned real interest rates in the inflation/deflation context. Inflation does not need to discourage savings as long as the saver is getting a high enough (real) interest rate. Actually, Scott keeps pointing this out. What does discourage savings is a situation like now, where I live, with 0.25% – odd interest p.a. on savings and 5% headline inflation.

    Symmetrically of course, deflation does not need to discourage lending as long as the debtor is paying a low enough (real) interest rate.

  43. Gravatar of Britmouse Britmouse
    15. December 2011 at 05:01

    http://www.bankofcanada.ca/2011/12/speeches/growth-in-the-age-of-deleveraging/

    Via FT Alphaville – here’s Mark Carney, another price stability junkie:

    Some have suggested that higher inflation may be a way out from the burden of excessive debt.

    This is a siren call. Moving opportunistically to a higher inflation target would risk unmooring inflation expectations and destroying the hard-won gains of price stability. Similarly, strategies such as nominal GDP level targeting would fail unless they are well understood by the public and the central bank is highly credible.

  44. Gravatar of Lars Christensen Lars Christensen
    15. December 2011 at 06:06

    Scott, this is quite interesting. I am sure George Selgin (until recently….) never heard of David Eagle. The same goes for David Eagle – he never knew about George Selgin (until recently). (And I never heard about David Eagle until recently…)

    Anyway, George of course in Less than Zero tells the correct story about good and bad deflation, but does not really stress in the same way the good inflation (though he does mention it…). I was there happy to see a paper by David Eagle recently where he stressed that there is good and bad inflation. His arguments where exactly the same as George’s (in a slightly different framework).

    The difference is the while George wrote “Less than Zero” when the latest policy mistakes was too high NGDP growth – David wrote his paper while the NGDP growth was too low.

    Obviously if you walk around with Selgin-Sumner-Eagle model in your head (I believe I do…) then you will not make the mistakes that you describe in your post…

  45. Gravatar of W. Peden W. Peden
    15. December 2011 at 06:24

    Lars Christensen,

    I think all these approaches are just the logical macroeconomic extension of Hayek’s work on prices: prices are signals and it’s a good idea to avoid distorting these signals. “Good deflation”, “Good inflation” etc. are the operation of supply and demand determining prices in a textbook microeconomic way, except at the aggregate level.

    That was what ultimately won me over to the idea of a gradually falling price level proposed by Selgin: in an economy with steady long-term increases in total factor productivity, there should be steadily falling prices. Otherwise, price signals are being distorted over the long run.

  46. Gravatar of John John
    15. December 2011 at 06:30

    Selgin,

    It’s John not Jon.

    I consider organic market processes to be “good” because they represent people’s freely expressed preferences. I consider interventions into the economy “bad” because they use coercion and coercion always has a loser while unhampered trades are a gain for both parties. In terms of stopping inflation or deflation, the only thing I would consider bad and worthwhile for the government to stop is their continual increases in the supply of money. That’s the cause of the vast majority of inflation as Friedman himself pointed out. Changes in demand for money rarely begin an inflation.

  47. Gravatar of Floccina Floccina
    15. December 2011 at 07:35

    (To Anybody)
    So was the housing bubble partly due to the fact that the fed was targeting CPI inflation in a period when China and India’s growth was causing real prices to fall? To keep inflation up did the fed push and hold interest rates too low, which in turn caused some to buy too much house because they thought that they would go up in value. It seems to dangerous when people start to buy things just because they think the future real value will be higher.

  48. Gravatar of Silas Barta Silas Barta
    15. December 2011 at 10:32

    Okay then, Scott_Sumner, could you at least stop feigning confusion on what kinds of inflation are bad, in light of my example? It’s fine if you want to curse the darkness, but I’ve got a match, like, right here and everything.

  49. Gravatar of Bill Woolsey Bill Woolsey
    15. December 2011 at 13:09

    I favor having nominal GDP grow at the trend growth rate of potential output. If it works perfectly, so nominal GDP stays on the target growth path, in deflation occurs with productive capacity grows more than trend. I have adopted the “good deflation” terminology.

    On the other hand, inflation would occur when productive capacity grows less than trend. When productive capacity actually falls, then it would be really high inflation. (More than 3% anyway.) I am not comfortable in calling this “good inflation.” It is a sign of a bad thing.

    Of course, I think that a monetary regime that caused spending to change in a way that avoided inflation or deflation in these situations would do more harm than good. Perhaps it is best just to give up on the “good deflation” terminology because the “good inflation” really doesn’t work.

    A monetary order that requires changes in the price level to adjust the real quantity of money to the demand to hold it is a bad thing. However, a monetary regime that forces the quantity of money away from the demand for money to force price changes is also a bad thing.

    Bartas:

    Outside of Hoppe (and maybe Harless) no one is arguing that the people hold money for too long or too short a period of time. They should be able to hold it like they want, and the quantity should adjust along with matching loans.

    Harless:

    Most money is interest bearing deposits. Why can’t it be added up? You seemed to be making an argument about fixed quantity, zero interest, outside money. I agree that there is a problem with a low price level creating greater real balances, greater real wealth, and reduced saving. I first saw the argument in Leijonhufvud. The solution is a different kind of money, I think.

  50. Gravatar of Bill Woolsey Bill Woolsey
    15. December 2011 at 13:21

    Flocina,

    I don’t really think that India and China are at issue.

    During the period in question, spending stayed on a 5% growth path. If we had a series of extra rapid increases in real output, so real GDP was growing 4%, 5% and we were having stable prices or 1% inflation, and the Fed sped up nominal GDP growth to get the price level back up, or at least to keep inflation at 2%, then that would have been the problem. But that isn’t what was happening.

    Look at the growth path of nominal GDP.

    In my view, CPI targeting could be worrisome. Suppose the GDP deflator is rising because new home prices and capital goods prices are rising. The CPI, on the other hand, isn’t rising. It would look to me like the demand for output was growing too fast, just not consumer goods. But that really isn’t what happened either.

    But what if consumer prices are kept from rising fast because the Chinese and Indians are sending savings to the U.S., which are used to fund to purchase of new housing. Resources are freed from producing consumer goods in the U.S. and instead used to build new houses.

    This is exactly what should happen. And only if it turns out that the houses were bad investments does it turn out to be a mistake in retrospect. And the Chinese and Indian investors should take a loss. Oh, if the U.S. government bails them out, then this whole deal looks pretty bad from the point of view of the U.S. taxpayer. But that doesn’t mean that the downward pressure on consumer goods prices offset by rising home prices was somehow wrong. That somehow a monetary contraction should have led to slower growth in demand for U.S. output so consumer goods would have fallen by more.

  51. Gravatar of W. Peden W. Peden
    15. December 2011 at 13:34

    Floccina,

    Partly perhaps (had the Fed been targeting <5% NGDP, the housing bubble wouldn't have been so bloated) but I think that the US case seems to have been more a matter of distorted relative prices between assets and consumer goods. I'm not sure that PT (i.e. NGDP + intermediate transactions + asset price transactions) can really be said to be out of control from 2001-2007, like it was in the late 1980s; a lot of the strength in property seems to have reflected a weakness in stocks e.g. the Dow Jones IA didn't reach its 2000 level until 2006.

    2006-2007 does seem to have been a time of significant overheating if you look at both NGDP and asset prices, but by then the US housing bubble was on the slide as far as I know. Unfortunately, I haven't seen anyone develop a really good proxy for US PT now that the ratio of PY to PT is (I assume) so unstable.

    It is perhaps audacious to mention it in a MM context, but it wasn't until 2006-2007 that M3 and M2 (and more importantly, the EXPECTED supply of money) started to really pick up. Both are flawed measures of broad money (M3 is too broad, M2 is too narrow) but both suggest that the US wasn't in an asset price/money supply spiral until 2006-2007 and of course this was quickly popped in 2008 when asset prices and broad money growth peaked.

    For all the importance of NGDP, I think that it's a bad guide to housing bubbles because bubbles have to be understood with reference to (1) the nominal wealth (not output) of a country, (2) the structural distribution of that wealth, and (3) people's preferences for their wealth portfolio. None of these are picked up by NGDP; what we do get from NGDP is a guide to what is happening to expenditure, which is good enough for monetary policy.

  52. Gravatar of W. Peden W. Peden
    15. December 2011 at 13:39

    Looking at the DIJA at bit: it’s interesting how the DIJA has nearly reached its 2008 value and may well have done had it not been for the European crises in the second half of this year. While the income velocity of M2 has fallen sharply this year, the transactions velocity doesn’t seem to have fallen so sharply (though the DIJA doesn’t tell the whole story, since the US property market is still in decline IIRC) and the DIJA has led M2 pretty well from 2008-2011.

    (M3 meanwhile continues to be a lousy indicator for NGDP but still gives a pretty good sense of what’s going on in the financial industry.)

    Having reddened Scott Sumner’s face with all this aggregate talk, I’ll stop.

  53. Gravatar of George Selgin George Selgin
    15. December 2011 at 18:16

    Pardon the typo, John…

    As for good inflation, under a productivity norm, the rate of inflation is the negative of the TFP growth rate; consequently “good” inflation (that is, inflation that is best tolerated) is relatively unusual. That’s the only reason I didn’t make much of it in Less Than Zero. I did and do nevertheless agree that if there’s a productivity setback you absolutely want prices to rise; indeed, price-level stabilization would be especially harmful in this case, for it would require deliberate contraction of NGDP.

  54. Gravatar of Jason Odegaard Jason Odegaard
    15. December 2011 at 18:48

    @Silas_Barta

    Scott was not expressing confusion about what type of inflation is good or bad. That was not the point of his post. He was asking why it seems impossible for some economists to accept the notion of good inflation. That was the question.

    As for the concept of money, getting back to the preference of non-monetary assets versus monetary assets.

    People *do* value money even though it *does not* provide indefinite value. I cannot think of any form of money that permits that capability. Generally money works in the short and medium term for that purpose. But not forever.

    What I am getting at, and I think Lorenzo is as well (though I don’t want to put words in his mouth), is the point that why should monetary policy prefer deflation that favors holding of monetary assets (cash and demand deposits) rather than non-monetary assets (bonds, stocks, land, commodities, etc). This reminds me quite a lot of this post by Nick Rowe which I thought sums it up well: http://worthwhile.typepad.com/worthwhile_canadian_initi/2011/08/recessions-are-always-and-everywhere-a-monetary-phenomena.html

  55. Gravatar of Silas Barta Silas Barta
    15. December 2011 at 19:54

    @Jason_Odegaard:

    Scott was not expressing confusion about what type of inflation is good or bad. That was not the point of his post. He was asking why it seems impossible for some economists to accept the notion of good inflation. That was the question.

    My response applies equally well as an answer to that question, so I don’t understand why you bring this up.

    People *do* value money even though it *does not* provide indefinite value.

    Indeed, but the justifications I see being given for sodomizing savers are all predicated on the inefficiency of holding on to money indefinitely, and therefore assume away a fundamental problem in proposing a solution.

  56. Gravatar of Jason Odegaard Jason Odegaard
    16. December 2011 at 06:41

    @Silas_Barta (but it’s just here now, I think) I brought it up since your posts and the point I think Scott was asking seemed to be talking past each other.

    I don’t intend to sodomize savers – I’d prefer they invest. Holding anything indefinitely is unpractical. The name of this blog is The Money Illusion. Perceiving money as an asset holding value the same way that aluminum, land, microchips, or trucks hold value is mistaken. That’s the point I’m trying to get across.

  57. Gravatar of Becky Hargrove Becky Hargrove
    16. December 2011 at 11:19

    Silas,
    Think of money as a train that goes down a single track. Hardly anyone ever leaves the train because it is practically the only way to reach the destination efficiently, at least in the present (walking or barter is slower). However, that means that the train continually picks up someone at numerous stops (societal expectations). So the train does not stay the same weight and has to burn more fuel just to maintain speed. How then is the train going to remain exactly the same?

  58. Gravatar of ssumner ssumner
    16. December 2011 at 11:55

    Silas, You said;

    “a) If inflation coincides with a booming economy, and people are better able to find their comparative advantages and thus employ themselves producing valuable things, that can outweigh the badness of inflation.”

    You have it backwards. It’s better for inflation to coincide with a weak economy, and deflation to coincide with a booming economy.

    John Thacker, I disagree. Conservatives think the natural disaster is a bad thing, and the rise in the price of water is a good thing which makes the natural disaster less bad. You need to think in ceteris paribus terms.

    Neal, And because I own stocks.

    Benny, I had thought that on most issues (other than econ) libertarians were even to the left of liberals.

    Rahul, Check out the items under “quick intro to my views” in the right column.

    mbk, Good point, although there is a tax distortion in our non-indexed taxes on interest and cap gains.

    Britmouse, He’s wrong, the public doesn’t have a clue as to how inflation targeting works. They were outraged when Bernanke announced inflation was too low and he was trying to raise their cost of living.

    Lars, I agree that David Eagle’s paper is very good.

    John, So are laws againsty counterfeiting “unjustified intervention in the free economy?” Why should the governemnt have a monopoly on M? And if they do, what makes a zero percent increase the optimal growth rate of government M? Was the gold standard to interventionist for you?

    Floccina, The Fed didn’t cause the housing bubble, the GSEs, TBTF, FDIC, etc, and tight money in late 2008 caused the housing bubble. And if those insurance funds must exist, then a lack of regulation caused the housing bubble.

    Bill, I agree.

    W. Peden, Don’t forget that stocks are being helped by foreign earnings in places like China.

  59. Gravatar of W. Peden W. Peden
    16. December 2011 at 12:06

    Scott,

    Hmm, that helps explain the divergence of PT and PY ratios. I read a paper recently that traced the instability of the relative transaction and income velocities of M4 to about 1976 (an awkward period because of the Corset distorting broad money aggregates, but there can be no doubt that after 1981 the divergence between PT and PY in the UK was dramatic). That matches up well with foreign earnings changing asset prices, since the large external capital flows from the UK kick off in 1981.

    The ratio stabilises a bit in the early 1990s, IIRC, before taking off in the late 1990s. That matches up with the decline of the Japanese economy, followed by the take-off of China and other non-Japanese external markets as a location for foreign investment.

    I did wonder a bit about whether it would be better to help get financial stability via targeting broad money (as a proxy for PT) or do NGDP targeting. I concluded that NGDP targeting was far more preferable, since we can get financial stability through reforming finance while keeping NGDP stable, whereas we can’t get NGDP stability while keeping broad money stable (not indefinitely anyway).

  60. Gravatar of W. Peden W. Peden
    16. December 2011 at 12:10

    (Although, with a significantly less-than-independent monetary authority such as Britain had until 1997, perhaps a K% type rule would be better than NGDP targeting* since estimating velocity would be a way that governments could game the system towards the election cycle. “We anticipate a fall in velocity over the next 36 months, which coincidentally has the election two-thirds of the way through it…”

    A K% rule has the advantage of being knave-proof, like the Gold Standard. Unfortunately, when income velocity undergoes big shifts, it is also stability-proof!)

    * EXCEPT a Market Monetarist NGDP target based on targeting market forecasts, which circumvents the political problem even in countries with politicised monetary authorities. 😉

  61. Gravatar of Silas Barta Silas Barta
    16. December 2011 at 12:33

    @ssumner: You have it backwards. It’s better for inflation to coincide with a weak economy, and deflation to coincide with a booming economy.

    *sigh* We’re not making any progress here. I’m trying to explain the senses in which inflation is bad, so as to resolve your feigned confusion. And I just went over, several times, why inflation *in and of itself* is bad, and why other coincident things can still be better than it. That means (much as you seem to be unable to comprehend) that you have to judge inflation’s goodness separately from the goodness of the other effects to see if the net effect is good.

    So your statement, in addition to being wrong, is made seemingly without having read anything I’ve posted here. It is wrong because inflation is always bad, whether in a booming or weak economy, and it is particularly bad when it’s making goods more expensive right while people have to draw down meagre savings!

    What you should be arguing, in order to make sense, is that in a weak economy, deliberately-engineered inflation is more likely to result in *other things* whose goodness outweighs the inflation’s badness. But alas, that would require you to be able to distinguish between an ends (inflation) and a means (prosperity), which you have time and again proven yourself to be utterly unable to.

    @Becky_Hargrove: Er, what? Money is like a train that doesn’t stop, so the savers are like added weight …? If you don’t understand analogies, please, just don’t use them.

    @Jason_Odegaard: Deferring consumption until you find something worth consuming is always and eminently wise. Policies that attempt to get people to consume earlier than they would like to on the grounds that “market exchanges are good, man!” are inevitably misguided because they fail to recognize that the purpose of economic policy is to enhance welfare, not goose semi-related econometrics like M*V.

    Purchases are not good. Good purchases are good. Purchases that are made solely to get the least bad outcome while money is withering away, beyond your control, are not good, and are not indicative of the kind of welfare enhancement that comes from freely-chosen purchases.

    If we could judge an economy “good” merely on the basis that there are lots of (necessarily Pareto-improving) exchanges, then we should be rooting for Weimar style environments, where workers must-must-must spend their money the moment they earn it — because after all, it’s like, inefficient and stuff, when people hold on to their money until they actually find something they want.

  62. Gravatar of A method to decompose supply and demand inflation « The Market Monetarist A method to decompose supply and demand inflation « The Market Monetarist
    17. December 2011 at 07:12

    […] and bad inflation. (For a discussion of this see Scott Sumner’s and David Beckworth’s posts here and here). Basically one can say that bad inflation/deflation is a result of demand shocks, while […]

  63. Gravatar of ssumner ssumner
    17. December 2011 at 08:56

    W. Peden, Don’t forget that PT is meaningless, as transactions in the financial sector now completely dominate goods transactions. Thus PT is totally disconnected from the real economy, and 100s of times bigger.

    Silas, Never reason from a price change. When prices rise, someone pays more and someone gets more. That’s a net wash. All that matters is how inflation affects the real economy; resource allocation, output, employment, consumption now, consumption in the future, etc.

    And there are some types of inflation that improve resource allocation and some types that make it worse.

    I’m amused that you think I ignored what you wrote. I provided a direct quotation, and then commented. I guess you are one of those “how dare you quote me” types of people.

  64. Gravatar of W. Peden W. Peden
    17. December 2011 at 15:45

    Scott,

    I don’t think that PT is meaningless just because it includes financial sector transactions. As we’ve found out in the past three years, financial sector transactions matter a lot, even if the link between PT and PY has broken down.

    You could say that it’s meaningless because it’s unmeasurable, but that’s overstating things: it’s very hard to measure and can’t be measured with total precision (neither can NGDP) but we can use proxies to get a sense of what’s generally going on with PT, just like we can use proxies for wealth.

  65. Gravatar of Scott Sumner Scott Sumner
    18. December 2011 at 08:49

    W. Peden, PT may matter in a micro sense, as in the many bad loans made in the subprime bubble. But those don’t drive the macro aggeregate PT. I’d guess daily transactions in forex markets dwarfs yearly totals of subprime loans.

  66. Gravatar of W. Peden W. Peden
    18. December 2011 at 08:59

    Both forex and bad loans matter. While the subprime mortgage fiasco didn’t make the Great Recession inevitable, it was surely an early spark.

    Also, if broad money is (1) endogenous and (2) can have causal impacts on PY, then surely PT is important because the prices of people’s assets has a big impact on their capacity to raise credit? Monetary policy matters a lot precisely because central banks can radically change the portfolios of economic agents, which are the P in PT.

  67. Gravatar of Scott Sumner Scott Sumner
    19. December 2011 at 14:44

    W. Peden, I’m not saying they don’t matter. I’m saying that PT is only very lossely correlated with the price of NGDP futures contracts (if they existed) and that we should ignore PT and focus on NGDP expectations. If the contracts don’t exist, I’d rather focus on equity and commodity prices, TIPS spreads, etc. than PT.

  68. Gravatar of W. Peden W. Peden
    19. December 2011 at 15:31

    Scott Sumner,

    “that we should ignore PT and focus on NGDP expectations. If the contracts don’t exist, I’d rather focus on equity and commodity prices, TIPS spreads, etc. than PT.”

    I quite agree. My point is that you can still have a lot of economic stability even if NGDP is stabilised. The solution is not to have the central bank target PT or adopt a K% rule, but rather to develop a financial sector which can react to changes in monetary policy in a more stable and sustainable way i.e. avoiding the sub-prime fiasco. Eliminating the moral hazard that comes from a system with super-profits and no real bankrupcy would be a salient step.

    In other words, since we can’t expect macro policy to provide financial systemic stability, we need a better micro policy. That is completely consistent with targeting NGDP.

  69. Gravatar of ssumner ssumner
    20. December 2011 at 10:10

    W. Peden, I agree.

  70. Gravatar of Free Banking » Good deflation and good inflation Free Banking » Good deflation and good inflation
    31. December 2011 at 20:16

    […] couple of weeks ago, Scott Sumner pointed out that many conservative-leaning economists think that certain types of deflation can be […]

  71. Gravatar of Andy Harless Andy Harless
    1. January 2012 at 19:41

    @Silas (if you’re still around)

    If it’s somehow sending an important market signal “” and not in any way introducing inefficiencies “” for me to hold my money a day before finding something worthy of purchasing, then why does this not hold for two days? Three days? A week, month, year, decade?

    Even short-term money holdings introduce inefficiencies (if there are sticky prices and a binding zero constraint on interest rates), but they are so small that they don’t matter. The issue is to make a tradeoff between the usefulness of money as a medium of exchange and the inefficiency introduced by sticky prices. There is some optimal level of money balances at which marginal cost equals marginal benefit.

    People would not value money unless it permitted them to indefinitely defer deciding what they will redeem it for.

    Yes they would. Suppose that, at the end of each month, all old money became invalid and had to be exchanged for government bonds or expire worthless, and then a new set of money were issued the next day and used via open market operations to purchase government bonds. In that case money would not permit people to indefinitely defer deciding what they would redeem it for: they would be required to redeem any remaining balances for government bonds at the end of each month. But presumably people would still value money, because (1) it could be used during the month to defer purchasing decisions and (2) any remainder could be redeemed for something valuable at the end of the month.

  72. Gravatar of Good deflation and good inflation – Alt-M Good deflation and good inflation - Alt-M
    29. March 2015 at 13:30

    […] couple of weeks ago, Scott Sumner pointed out that many conservative-leaning economists think that certain types of deflation can be […]

Leave a Reply