Aggregate demand–it’s not what you think

Most people have some understanding of “demand,” at the level of individual products.  Most people think aggregate demand is somehow related to that concept of demand.  It isn’t.

Here’s an interesting exchange between Matt Yglesias and Michael Mandel:

I find the “slow rate of innovation” hypothesis much more convincing than weak demand http://marginalrevolution.com/marginalrevolution/2013/11/are-real-rates-of-return-negative-is-the-natural-real-rate-of-return-negative.html …

@MichaelMandel For any given level of innovation, a level of demand that produces high unemployment and low inflation is inadequate.

@mattyglesias Innovation–the creation of new goods and services that people want to buy– is what produces higher level of demand

(Yglesias had the middle comment, Mandel the other two.)

At first I thought Mandel was making some sort of “Say’s Law” argument.  Supply creates it’s own demand. If so, they would have been talking past each other, as Matt Yglesias actually is talking about aggregate demand, whereas Say was talking about what we’d now call “quantity demanded.”  Say was saying that if the LRAS curve shifts to the right, there will be a new equilibrium at a higher level of quantity demanded, even if the AD curve does not shift at all.  And that’s true.  But of course it doesn’t address the claim of people like Matt and myself, which is that AD is too low, that the AD curve is too far to the left.   That is a problem, at least in the short run (Say’s Law fixes it in the long run.)

On second reading I wonder if I misread Mandel.  Perhaps he meant that companies needed to produce more nifty products, to whip up enthusiasm among consumers.  In other words, he’s saying there are too many products out there like BlackBerry, which are not innovative, and hence consumers are sitting on their wallets.

If so, that would also be wrong, but in an interesting way.  AD has nothing to do with “demand” in the microeconomic sense.  Matt’s right that in the 1930s there were plenty of nifty products; the problem was that consumers had too little money to buy the goods.  What would it look like if lack of nifty products really were the problem?  Imagine next Tuesday every store in American announced a sale; every product was priced at one cent.  Even Tiffany diamonds, Ferraris, Hollywood homes, etc.  Also assume that shopping did not pick up on that day.  Then we’d be living in a Mandel world, were a lack of nifty products was holding back AD. Of course we clearly do not live in that world.

At the micro level “demand” is a sort of real concept, the amount of BlackBerries that consumers want to buy at various prices.  Aggregate demand is nothing like that.  AD went up more than a trillion-fold in Zimbabwe a few years back, and yet “demand” as most people visualize the term remained anemic.  (I.e. demand in terms of relative prices.)  AD is a nominal concept, a concept related to money, not consumer goods.  Only the central bank can solve our AD problem, no one else has the proper tools.

If consumers have the money, I assure you there will definitely be things they want to buy.  I’m already preparing my shopping list just in case all stores announce a one cent sale next Tuesday.

PS.  Or maybe Mandel had a third meaning that I missed, in which case—never mind.

PPS.  Counterarguments based on income inequality won’t help either.  That’s simply a one-time velocity shift; monetary policy still drives AD.


Tags:

 
 
 

59 Responses to “Aggregate demand–it’s not what you think”

  1. Gravatar of Adam Adam
    20. November 2013 at 15:03

    It seemed clear to me that Mandel had no understanding of what aggregate demand is. But it’s a common error. “If stuff was better, people would want more stuff,” makes intuitive sense, and is probably true. It just has nothing to do with whether the macroeconomy has an aggregate demand problem.

  2. Gravatar of jknarr jknarr
    20. November 2013 at 15:37

    Smoke and mirrors, mostly. There are system-wide effects to the aggregate demand argument side: the reason that you are able to take advantage of the wonderful technology of jet flight 20 days of the a year is because there is a immense pool of demand that keeps flights running the other 345 days.

    Demand is what ensures systematic widespread adoption — and sometimes locks in suboptimal technologies (qwerty). There is technological retrograde, too. The Antikythera mechanism was never adopted, although the techology existed. Chinese treasure ships came and went.

    There’s plenty of innovation: our world could be wired up to computers from the smallest- to largest- tasks. But, income/demand is not there to encourage mass adoption.

    Instead, technological innovations remain mostly in the hands of big balance sheet entities who use it to build Panopticon prisons of control, and to extract monetary value (before the citizenry gets a shot at adoption).

    In short, tight money policies – which result in low aggregate demand – has increased the technological spread between the citizenry and the oligarchical large balance sheets of society (wall street, DC).

  3. Gravatar of Mark A. Sadowski Mark A. Sadowski
    20. November 2013 at 16:35

    “Matt’s right that in the 1930s there were plenty of nifty products; the problem was that consumers had too little money to buy the goods.”

    Ironically, the consensus is that the US rate of technological innovation reached its peak from 1929 to 1941, or precisely during the dozen years that spanned the Great Depression.

    http://lsb.scu.edu/economics/faculty/afield/AER%20September%202003e.pdf

    The Most Technologically Progressive Decade of the Century
    Alexander J. Field
    September 2003

    Abstract:
    “There is now an emerging consensus that over the course of U.S. economic history, multifactor productivity grew fastest over a broad plateau between 1905 and 1966, and within that period, in the two decades following 1929. This paper argues that the bulk of the achieved productivity levels in 1948 had already been attained before full scale war mobilization in 1942. It was not principally the war that laid the foundation for postwar prosperity. It was technological progress across a broad frontier of the American economy during the 1930s.”

  4. Gravatar of Mike Freimuth Mike Freimuth
    20. November 2013 at 16:56

    I think most people who become Austrians do so because they get introductory micro but then run into the “aggregate demand” conundrum in macro and can’t reconcile it with micro demand. (I was one of these people.) It would be better if it had a different name. I don’t think very many people understand what aggregate demand really means. jknarr is illustrating this with a comment that has nothing to do with aggregate demand in spite of the pretty good explanation in this post.

  5. Gravatar of Geoff Geoff
    20. November 2013 at 17:20

    Recall THIS recent Money Illusion post which reminded us that:

    “[E]very transaction is two-sided. When someone buys something, someone else sells something. So when you describe the quantity of something purchased changing, you are also describing the quantity of something sold changing.

    Now look at Sumner’s argument in this post:

    “AD has nothing to do with “demand” in the microeconomic sense. Matt’s right that in the 1930s there were plenty of nifty products; the problem was that consumers had too little money to buy the goods.”

    According to earlier Sumner, this argument is the same argument as “Sellers had too few goods to buy the money.”

    But what does Sumner argue, when the implication is against “demand management”? He tries to show it doesn’t work after all via an attempt at a reductio ad absurdum argument:

    “What would it look like if lack of nifty products really were the problem? Imagine next Tuesday every store in American announced a sale; every product was priced at one cent. Even Tiffany diamonds, Ferraris, Hollywood homes, etc. Also assume that shopping did not pick up on that day. Then we’d be living in a Mandel world, were a lack of nifty products was holding back AD. Of course we clearly do not live in that world.”

    Notice how this doesn’t even postulate a world where there are no, or fewer goods available for sale that buyers wanted. Here, the assumption is different, indeed opposite: We’re supposed to imagine plenty of supply. How is that a
    lack of goods buyers want to buy” scenario?

    What happened during the early 1930s is that buyers AND sellers were not willing to trade goods for money in a similar way as they did prior. The claim that goods were not sold because buyers did not have enough money…and that’s it, end of story, is not only incomplete…according to the logic Sumner used just 11 days ago, but it is also wrong on its face. Money existed in the early 1930s. The question that must be answered is why did so much of that money not get spent? Why did people decide to “hoard” a substantial quantity of money, instead of spending it on goods? Or, and equivalently, why did sellers decide to “hoard” a substantial quantity of goods, instead of spending it on money?

    To answer based solely on “demand” fails to take into account the reasons why sellers did not have enough goods to sell for the money prices that were being offered. The argument that there was not enough money spending is, as Sumner noted 11 days ago, equivalent to there not being enough goods spending.

    If the response to this is that “The goods were available to be sold! The owners of goods just weren’t willing to sell at the prevailing bid-ask exchange ratios!” can be met with the equivalent argument that “The money was available to be sold! The owners of money just weren’t willing to sell at the prevailing bid-ask exchange ratios!”

    Market monetarism, by its very construction, necessarily puts blinders on people to focus on one side of the equivalence coin, and as a result, ignore or become confused over the other side of the coin.

  6. Gravatar of Geoff Geoff
    20. November 2013 at 17:44

    “If consumers have the money, I assure you there will definitely be things they want to buy. I’m already preparing my shopping list just in case all stores announce a one cent sale next Tuesday.”

    This comment shows a lack of understanding market processes. The goods that exist are a result of past productive decisions that were founded on a particular relative price structure. It is not a valid thought experiment to imagine all of a sudden all goods are available for $0.01, if the intent is to replicate a problem on the real side. Production requires investment spread out across various stages. The relative prices that result prevents a thought experiment of $0.01 goods from staying true to the premises of economic activity in a division of labor.

    The idea that because we can imagine lots of goods being sold for $0.01, it means that more money really is a solution, can be shown as absurd by simply considering the equivalent side of the exchange coin, namely:

    “If sellers have the goods, I assure you there will definitely be things they want to sell. I’m already preparing my selling list just in case all buyers announce a one cent sale next Tuesday.”

    ——————

    If the problem is that the wrong goods were produced, or not enough of the goods that people want and too much of what they don’t want any more of, given available resources, then more money cannot solve this problem. More money cannot force people to clear a market that has not enough of some things (which we can’t see) and too much of other things (which we can see). For in this case, more money, if it is going to have a “positive effect” on output as it taking place, must be allocated in such a way that relative spending and relative prices remain on the same trajectory, which of course will just encourage and exacerbate the production problems on the real side!

    And, very importantly, a scenario of too many goods of one sort being produced, and not enough goods of another sort being produced, looks EXACTLY like a “general over-production” and/or “insufficient demand/spending”! For when too many of some goods are produced, and not enough of other goods, the result is excess inventory, and rising cash holding times as people wait for the right goods to be produced!

    Sumner is incapable of understanding how more money even within his own desired constraints would hamper rather than promote production. He refuses to accept that production problems can be significant due to constant growth NGDP, due to its effect on relative spending and relative prices, and thus on allocation of scarce resources, which has the result of partial relative over production and partial relative underproduction. He looks at the results of this, and he concludes not enough money/spending, when the real problem is structural. He can’t understand it because the portion of production that is unsufficient, we can’t see, because those goods were not produced. So Sumner’s bias towards positivism leads him to rejecting a reality that can only be known via understanding, not observation.

  7. Gravatar of Gordon Gordon
    20. November 2013 at 18:26

    Thanks Scott. This post was very informative. As a non-economist, I’ve been wrapping my head around MM by focusing on MV=PY. I had taken a glance at some AD-AS diagrams but had not really made the effort to understand them or read anything to explain what they meant. Now I think I understand what the AD curve means and why there is a SRAS curve. I agree with Mike that the term “aggregate demand” is misleading. On an AD-AS diagram, why not call the AD curve the NGDP curve instead?

  8. Gravatar of Michael Michael
    20. November 2013 at 18:35

    Geoff wrote:

    “And, very importantly, a scenario of too many goods of one sort being produced, and not enough goods of another sort being produced, looks EXACTLY like a “general over-production” and/or “insufficient demand/spending”! For when too many of some goods are produced, and not enough of other goods, the result is excess inventory, and rising cash holding times as people wait for the right goods to be produced!”

    No. If too many goods of one sort were produced (with a resultant shortage of other goods) relative prices would adjust such that markets were cleared.

  9. Gravatar of Mike T Mike T
    20. November 2013 at 18:58

    Scott,

    “At the micro level “demand” is a sort of real concept, the amount of BlackBerries that consumers want to buy at various prices. Aggregate demand is nothing like that. AD went up more than a trillion-fold in Zimbabwe a few years back, and yet “demand” as most people visualize the term remained anemic. (I.e. demand in terms of relative prices.) AD is a nominal concept, a concept related to money, not consumer goods. Only the central bank can solve our AD problem, no one else has the proper tools.”

    >>If real demand only has meaning at the micro level in terms of “relative prices,” while nominal demand in the aggregate is unrelated to consumer goods and misleading, at best, from the perspective of an individual’s well-being, then why is AD a “problem” the central bank must solve?

  10. Gravatar of Mike Freimuth Mike Freimuth
    20. November 2013 at 19:11

    Geoff,

    You said:

    “The question that must be answered is why did so much of that money not get spent? Why did people decide to “hoard” a substantial quantity of money, instead of spending it on goods? Or, and equivalently, why did sellers decide to “hoard” a substantial quantity of goods, instead of spending it on money?”

    That’s true but you are acting like no attempt has been made to answer this question. Scott is constantly attempting to explain what causes changes in aggregate demand. In this particular post he is just trying to explain what is meant by the term “aggregate demand” because many people (including you judging from these comments) don’t really understand it. Instead of trying to figure out what people who talk about aggregate demand mean, you’re just giving a canned rant against everything you associate with people who talk about aggregate demand.

    The term he is trying to explain is essentially just a name for the thing you correctly identify as the essential question underlying all of this. How are we to answer that question without giving that thing a name? You seem to just want everyone who doesn’t subscribe to your model to abandon all the concepts and terminology that don’t conform to it. I don’t see how you think you are advancing any method of thinking by doing this.

  11. Gravatar of Mike Freimuth Mike Freimuth
    20. November 2013 at 19:16

    Geoff,

    And by the way, quantity purchased/sold is different from demand (or supply) that basically the whole point of this post and the part of a previous post you are quoting. These two points are not contradictory in any way. Also the other Michael is right about relative prices changing assuming that prices aren’t sticky. That’s the weird thing about your arguments. All Keynesian and MM arguments rely on price stickiness for most of their implications. You seem half the time to implicitly deny that prices can be sticky but then still arrive at conclusions that are only possible if prices are sticky without ever actually addressing this question or recognizing that it is important.

  12. Gravatar of Mike Freimuth Mike Freimuth
    20. November 2013 at 19:21

    Mike T,

    I think you are misreading the post. It’s not that AD is “unrelated to consumer goods and misleading, at best, from the perspective of an individual’s well-being.” It’s just not anything like an actual demand curve. It’s still relevant (Scott would say of primary importance) to monetary policy which has real implications, “at least in the short run.”

  13. Gravatar of Geoff Geoff
    20. November 2013 at 19:43

    Michael:

    “No. If too many goods of one sort were produced (with a resultant shortage of other goods) relative prices would adjust such that markets were cleared.”

    You need to be more careful Michael. Yes, in an unhampered market, relative prices will adjust and that will tend to reverse the partial relative over and underproductions.

    But the Fed keeps reversing this because relative price corrections require not only changed relative spending, but changed production as well. Mere price changes cannot by themselves solve the problem that needs changed production.

    Relative price changes cannot immediately solve the problem of partial relative over production and partial relative underproduction, for the partial relatively underproduced goods do not exist as of yet. Prices, as I am sure you know, can only exist for goods that do exist. And, what’s more, it’s not always the case that the underproduced goods are merely goods that do exist but are just in insufficient numbers.

    Relative price changes cannot solve the problem, if relative prices keep being hampered by central bank targeting of price levels or aggregate spending, for the way inflation transmission works is on the ground through relative spending and price changes. Sure, you can measure this in terms of aggregates, but there are relative changes that constitute the aggregate changes.

  14. Gravatar of Dustin Dustin
    20. November 2013 at 19:45

    Great post. I was asking Mark A. Sadowski on another blog recently to clarify what AD was, it was hard for me to see as more than the sum of quantities demanded.

  15. Gravatar of Geoff Geoff
    20. November 2013 at 19:52

    Mike Freimuth:

    “That’s true but you are acting like no attempt has been made to answer this question. Scott is constantly attempting to explain what causes changes in aggregate demand. In this particular post he is just trying to explain what is meant by the term “aggregate demand” because many people (including you judging from these comments) don’t really understand it. Instead of trying to figure out what people who talk about aggregate demand mean, you’re just giving a canned rant against everything you associate with people who talk about aggregate demand.”

    There are a number of problems with this comment. First, it is likely I understand aggregate demand more than you or Sumner, judging by your comments. For example, in the aggregate, I notice him and others fail to show an understanding that spending on final goods is actually in competition with all other spending, including wage payments. Also, you tend to take a holistic approach abstracted from the driver of AD, whereas I keep the driver in view at all times. Secondly, I am not trying to figure out what Sumner means by aggregate demand in this post, but rather I am showing that according to Sumner’s past posts, he is making an inconsistent argument. Thirdly, I am not ranting against anyone who talks about aggregate demand. There is nothing wrong with talking about aggregate demand.

    “The term he is trying to explain is essentially just a name for the thing you correctly identify as the essential question underlying all of this. How are we to answer that question without giving that thing a name? You seem to just want everyone who doesn’t subscribe to your model to abandon all the concepts and terminology that don’t conform to it. I don’t see how you think you are advancing any method of thinking by doing this.”

    What is this “you seem” nonsense? How about we address what each other are actually saying? No, this is not a semantic debate. Semantics is usually the recourse for people such as yourself who have trouble engaging arguments that you don’t understand. Perhaps rereading what I wrote and not being biased like you are, will allow you to grasp the point this time?

  16. Gravatar of Benjamin Cole Benjamin Cole
    20. November 2013 at 19:55

    Mandel is just off.

    Besides, even if we buy into Mandel’s way of looking at the world, if consumers are bored at the latest gadgets, there are still vacations, hair salons (women get tired of new hairdos and clothes? Yeah? When?), duding up your house (you can sink real money there), expensive ballgames, expensive girlfriends, booze, country club memberships and so forth.

    Restaurants? People get bored of eating?

    Health care spending is whole ‘bother ballgame.

    We suffer from a lack of AD in the USA. Print more money is the immediate solution, and supply-side smarts is the long-run game.

  17. Gravatar of ssumner ssumner
    20. November 2013 at 19:55

    Adam, It’s always possible I misunderstood him.

    Mark, That’s right.

    Mike and Gordon, Yes, the term “aggregate demand” is really unfortunate. It has nothing to do with demand in a microeconomic sense.

    I’d prefer “nominal spending” or “nominal income”.

    MikeT, Because the real problem is too little NGDP, and that is a problem only the central bank can solve.

  18. Gravatar of Geoff Geoff
    20. November 2013 at 19:56

    Mike Freimuth:

    “And by the way, quantity purchased/sold is different from demand (or supply) that basically the whole point of this post and the part of a previous post you are quoting. These two points are not contradictory in any way.”

    Outstanding.

    “Also the other Michael is right about relative prices changing assuming that prices aren’t sticky. That’s the weird thing about your arguments. All Keynesian and MM arguments rely on price stickiness for most of their implications. You seem half the time to implicitly deny that prices can be sticky but then still arrive at conclusions that are only possible if prices are sticky without ever actually addressing this question or recognizing that it is important.”

    Relative prices cannot change to clear the market if the Fed keeps reversing these changes with opposite relative price changes that caused the partial over and underproduction from occurring in the first place.

    Price stickiness is a red herring actually. Regardless of how fast or slow individuals change prices, in response to demand and/or supply changes of other individuals, is not particularly relevant to what I am saying.

  19. Gravatar of Mike T Mike T
    20. November 2013 at 19:58

    Mike Freimuth,

    “I think you are misreading the post. It’s not that AD is “unrelated to consumer goods and misleading, at best, from the perspective of an individual’s well-being.” It’s just not anything like an actual demand curve. It’s still relevant (Scott would say of primary importance) to monetary policy which has real implications, “at least in the short run.”

    >> I get the post. I’m just trying to tease out what the ultimate goal is here. Is the “real” problem AD or sticky prices? Surely, even in a completely unhampered free market, mistakes will be made (or industries disrupted through innovation) and adjustments will be necessary which would require price changes that would take some positive time “t”. But if we’re debating ideas and economics, isn’t it more fruitful to attack those non market factors which exacerbate price stickiness (e.g. subsidies, bailouts, minimum wage laws, prolonged unemployment benefits, and a myriad of other government regulations) rather than succumbing to our current monetary and political arrangements choosing the path of least resistance since we have a Fed with discretionary monetary policy? In other words, is addressing AD through monetary policy primarily one of tactical convenience? And are you willing to at least entertain the possibility that monetary policy disrupts, rather than alleviates, necessary relative price adjustments?

  20. Gravatar of Geoff Geoff
    20. November 2013 at 20:05

    The problem isn’t too little NGDP. The problem is too much spending in some sectors and not enough spending in other sectors, which would tend to correct the real underlying problems.

    Sometimes, a falling NGDP is the only way for relative spending to change in a way that corrects partial relative over and underproduction, for the alternative of the Fed increasing NGDP would come at the cost of reversing the needed partial relative spending changes that are necessary to correct the actual problems.

    Remember, when the Fed affects NGDP, it is not raising every micro level spending act from the individual in exactly the same way. For example, if the Fed put NGDP growth from 4% to 5%, then it is not the case that every individual spends that much more on every purchase. What actually happens is that, at first, some individuals spend more money while others do not spend any more money, which has the effect of some goods experiencing more spending whilst others do not. NGDP is still rising, but relative spending is changing, and that affects production in a non-homogeneous manner.

    Problems caused by partial relative underproduction and partial relative overproduction APPEAR as insufficient aggregate spending, because the response by individuals to these real problems is to hold cash for longer while waiting for the relative production changes to take place. This of course results in a falling NGDP, ceteris paribus.

    The Fed fixing NGDP isn’t going to solve the problems that are caused by inflation itself. We are of course supposed to minimize these problems.

  21. Gravatar of Geoff Geoff
    20. November 2013 at 20:12

    Mike T:

    “But if we’re debating ideas and economics, isn’t it more fruitful to attack those non market factors which exacerbate price stickiness (e.g. subsidies, bailouts, minimum wage laws, prolonged unemployment benefits, and a myriad of other government regulations) rather than succumbing to our current monetary and political arrangements choosing the path of least resistance since we have a Fed with discretionary monetary policy?”

    Thank you for being among the rare breed on this blog who understands the most pragmatic course of action to take, given the world we live in.

    I would only add that price stickiness is itself exacerbated by inflation, the very cure for problems allegedly caused by price stickiness. This is because workers become psychologically adapted to continuously rising prices all the time, which of course makes them (and employers) very resistant to cutting wages drastically even though it would be to their interests. We have macroeconomists to blame as well, who have relentlessly preached to the public that they must fear price deflation. Sumner is guilty of this. So are many others.

    It’s an interesting sight to behold. We have to endure pundits make a mess of society, and then have to suffer even more as they insist they are the only ones who can fix the problem they themselves caused, and to top it all off, we have to pretend they’re not insane as they continue to repeat the solutions that are just more of the same problem causing activities.

  22. Gravatar of Geoff Geoff
    20. November 2013 at 20:17

    How will market monetarism solve the structural problems caused by relative spending changes brought about by non-market means such as the Fed purchasing assets from banks? It can’t. Going from price level targeting to aggregate spending targeting can’t do it, because it’s just a change in the quantity, not the transmission, of inflation.

  23. Gravatar of Geoff Geoff
    20. November 2013 at 20:38

    The Fed’s inflationary, bail out everyone response from 2008 on, has had a devastating effect on production. For now everyone believes in the Greenspan Put. There is no credible risk management any longer. This is dragging down growth and employment.

    But we’re supposed to believe it’s insufficient inflation. Ha.

  24. Gravatar of Mike T Mike T
    20. November 2013 at 20:58

    Geoff

    “I would only add that price stickiness is itself exacerbated by inflation, the very cure for problems allegedly caused by price stickiness.”

    >> Agreed. That’s why I slipped in “And are you willing to at least entertain the possibility that monetary policy disrupts, rather than alleviates, necessary relative price adjustments?” at the end of my comment as well. I was just attempting to approach it from a different angle.

    “We have to endure pundits make a mess of society, and then have to suffer even more as they insist they are the only ones who can fix the problem they themselves caused, and to top it all off, we have to pretend they’re not insane as they continue to repeat the solutions that are just more of the same problem causing activities.”

    >> You have to at least admit that it’s a shrewd self-preservation mechanism.

    I enjoy reading your comments here and they should be taken seriously.

  25. Gravatar of Saturos Saturos
    20. November 2013 at 21:12

    “consumers had too little money to buy the goods” sounds more like a statement about MoE than MoA… 😛

  26. Gravatar of Geoff Geoff
    20. November 2013 at 21:15

    Mike T:

    Uh oh, get ready for Sumner to have another intellectual meltdown, in trying to convince you (himself?) that I am the wrongiest wrongy wrong person ever. He can’t help himself. Grab your popcorn!

  27. Gravatar of Mike Freimuth Mike Freimuth
    20. November 2013 at 22:55

    Mike T,

    I actually agree with you that sticky prices are largely caused by government and we would be better off without the policies you mention. However, that is a separate issue from the meaning of aggregate demand (which is what the post is about). I think the following represents Sumner’s view but if not I will leave it to him to correct it.

    First, regarding this: “And are you willing to at least entertain the possibility that monetary policy disrupts, rather than alleviates, necessary relative price adjustments?”

    I am willing to entertain that possibility but I think the premise behind that question is flawed. The question is not what monetary policy does compared to the absence of monetary policy. The question is what a certain monetary policy does compared to a different monetary policy. In the current system, there is no such thing as no monetary policy. Scott thinks policy should be looser because he thinks that current policy is producing a level of AD which is below trend. Understanding why that matters requires one to understand what he means by AD which is the point of the post.

    Regarding the question whether the “real” problem is AD or sticky prices, it is AD but only because there are sticky prices. If prices were not sticky, monetary policy would be neutral even in the short run.

  28. Gravatar of Mike Freimuth Mike Freimuth
    20. November 2013 at 23:14

    Mike T,

    Also, I’m willing to consider it but I haven’t heard an explanation yet of how loose monetary policy (which is presumably what you mean by monetary policy) disrupts necessary relative price adjustments. What causes this?

    Wages are typically considered to be the main sticking point and these are often assumed to be only sticky in the downward direction. If they were sticky upward, perhaps through long-term contracts, one could argue that excessively loose monetary policy could have a similar (but opposite) effect on relative prices to excessively tight policy. However, I imagine that this is not what you have in mind. Can you give me another mechanism to consider?

  29. Gravatar of Morgan Warstler Morgan Warstler
    21. November 2013 at 00:24

    ARRGGHHH!

    Look guys, this is the EVERYTHING:

    http://www.asymco.com/wp-content/uploads/2013/11/Screen-Shot-2013-11-19-at-11-19-8.04.42-PM.png

    What have I always taught you?

    The WAY TO MEASURE ECONOMIC INEQUALITY IS WHAT?

    As a time series.

    HOW LONG does it take to disperse technology to the have nots?

    How many times have I gone on and on about that $50K 60″ HDTV becoming a $500 product?

    How many times have I talked about MOVEMENT TO FREE, and how it is possible but not healthy to escape the gravity of technology driven deflation?

    THAT CHART is telling you the longer it takes tech to go thru the system, the more GDP is an acceptable measure of humankind.

    And the reverse is true as well.

    Greenspan has this thing about measuring the actual WEIGHT of the economy.

    He is correct.

    I have a buddy who is CEO of a the market leader in abrasives, who does a rough sketch of basically all product prices based solely on their weight, and lowers that cost every year.

    We in tech, expect everything to cost “a dollar a pound.”

    Nobody is lying. Nobody is wrong.

    That chart tells you why Nominal GDP needs grow very slowly and get slower – like 4%, 3%, 2%, 1%…

    We are consume less weight. The real cost of weight is declining.

    Money Illusion may be real. But it existed in a time where it never had to deal with FREE.

  30. Gravatar of Morgan Warstler Morgan Warstler
    21. November 2013 at 00:40

    We all know there is no unemployment problem, GI / CYB prove it beyond shadow of doubt.

    Consumption amongst poor is higher than it has every been.

    Inequality is at its lowest levels in human history, the poor get new stuff faster than ever.

    These are facts.

    And we are headed into an age of abundance PRECISELY BECAUSE prices and incomes are going to feel immense downward pressures.

    MM and NGDPLT should be about Accepting but also confronting the Money Illusion.

    Because Money Illusion is a minor deity compared to the economic god Scarcity, and the digital free has toppled Scarcity, we need to tell our story with that in mind.

  31. Gravatar of Morgan Warstler Morgan Warstler
    21. November 2013 at 01:03

    Here’s Ryan Avnet graph:

    http://research.stlouisfed.org/fredgraph.png?g=ozp

    While consumption has exploded, the price level has headed toward “stagnant”

    It costs less money to start things, each new thing started reduces prices being paid, Money Illusion is standing on the tracks saying “but, but, but people think nominally!”

    Taking a bat to sticky wages and prices is crucial.

    A small level NGDP target that economic players can bank on is crucial, so we can swing the bat as hard and viciously as possible.

  32. Gravatar of Ravi Ravi
    21. November 2013 at 03:22

    Off-topic, Scott, I finally had a chance to listen to your presentation at Cato. Thought it went very well, and you’re to be commended for refining your presentation over the years, even to a hostile crowd. I hope the overwhelming number of questions aimed at you was an indication of general interest and not just sniping (although I have to say that supposed students of monetary matters should have been more familiar with your arguments. Either they ignore the blogosphere or have no intention of trying to pass Bryan Caplan’s ideological Turing test). In any case, congratulations, and thank you for your continued efforts to educate people.

  33. Gravatar of Thursday Morning Links | timiacono.com Thursday Morning Links | timiacono.com
    21. November 2013 at 03:38

    […] ‘Manly’ jobs aren’t coming back – CNN/Money Aggregate demand-it’s not what you think – Money Illusion Americans’ Gaping Hole in Their 3-Month Safety Net – Fiscal Times […]

  34. Gravatar of Matt McOsker Matt McOsker
    21. November 2013 at 05:57

    I don’t think it takes “nifty” products, but people need money to buy products. Look at Home Depot’s revenues in 2004 were $64.8 billion in 2007 it peaked at $90.8 billion. It dipped to a low of $66 billion in 2010, and is currently at $74.7 billion. That cycle was driven by money from the real estate boom. I am pretty sure Home Depot can access plenty of supply, but consumers really are strapped at the money end to buy that supply. I know this is just one company and industry, but IMO it tells you a lot of where things have been the past decade.

  35. Gravatar of TravisV TravisV
    21. November 2013 at 06:29

    Prof. Sumner,

    3.6%??? Doesn’t this seem way too optimistic to you?

    “Goldman Sachs is bullish on stocks. In a new note, chief equities analyst David Kostin predicts growth will lead U.S. markets as much as 23% higher through 2016. “Our [2014] return forecast reflects rising, albeit decelerating, profit growth and a slightly lower P/E multiple,” said Kostin. “The linchpin of our market forecast is growth – in the economy, sales, and earnings. We expect 3.6% global economic growth. The US will advance at a 3% pace while inflation remains contained at 1.4%. China, Japan, and even Europe will all grow, expanding GDP by 7.8%, 1.6%, and 1.5%, respectively.”

    Read more: http://www.businessinsider.com/opening-bell-november-21-2013-2013-11#ixzz2lI6j9SHa

  36. Gravatar of JohnB JohnB
    21. November 2013 at 06:46

    To me the real question is why the economy hasn’t been able to match full employment with the current level of AD? Why hasn’t Say’s Law kicked in faster? Just like in the Great Depression, I suspect two main causes.

    1. Activist government policy and regulatory changes creates an environment of uncertainty where private business doesn’t want to invest and hire new workers. Companies sitting on cash and banks holding so much in excess reserves are exhibits A and B.

    2. Wage rigidity created by the minimum wage (and it’s 2009 increase) along with things like extended unemployment insurance that effectively increase the price at which people are willing to work.

    The economy should be able to adjust to slower Nominal GDP growth over a 5 year period. In fact, the economy should be able to adjust to almost any shock to NGDP. That’s what makes it hard for me to buy Aggregate Demand arguments. Aggregate Demand may be able to explain one off events but not 5 or more years of no recovery. According to labor force participation, the best indicator and most important measure of whether the economy is working at capacity, 5 years of no recovery is exactly what we’ve had.

    P.S. The demographic arguments about labor force participation are bunk. It’s not like all the baby boomers spontaneously retired in 2008-2009. If you believe that, you’d also have to believe that the amount of people getting seriously injured on the job also spiked in 2008-2009.

  37. Gravatar of ssumner ssumner
    21. November 2013 at 06:59

    Saturos, Good catch, Of course I meant “monetary income.”

    Thanks Ravi.

    Travis, That might be slightly too high for global growth, but it’s not far off base. I might forecast 3.3%.

    JohnB, You are partly right about supply-side problems, but your facts are wrong. It’s not correct to say there has been no recovery; RGDP has been rising and unemployment has fallen from 10% to 7.3%. You also ignore the fact that the recovery in AD has been far slower than in previous recoveries, so you’d expect a slower recovery in jobs this time around.

  38. Gravatar of Saturos Saturos
    21. November 2013 at 07:16

    Yellen just confirmed by Senate.

  39. Gravatar of Andrew Andrew
    21. November 2013 at 07:48

    Venezuela is trying your “everything for a penny” idea right now.

  40. Gravatar of ssumner ssumner
    21. November 2013 at 08:16

    Andrew, Yes, and it’s “working.” 🙂

  41. Gravatar of Michael Michael
    21. November 2013 at 08:30

    NGDP has been growing at about 4% per year since the 2009 trough, correct.

    What would happen if the Fed announced 4% NDGPLT, effective tomorrow, but with no attempt at any catch up growth? Would that be good, bad, or irrelevant?

    In one sense, it would not represent a change in policy (we already have a 4% growth rate).

    On the other hand, if seen as credible, it would allow for better forward guidance and less taper-related chaos.

  42. Gravatar of Doug M Doug M
    21. November 2013 at 09:24

    If the price level is increasing but Real GDP is unchanged, unemployment is unchanged (and high), is Aggregate Demand increasing.

    Scott Sumner says Yes! By definition. I am not sure how much of the economics profession would agree. And, how much of the “laity”?

  43. Gravatar of Steven Kopits Steven Kopits
    21. November 2013 at 10:07

    Ah, what do Americans demand?

    Oil, it turns out. Demand is up 9.9% in the last four weeks compared to the same period last year, according to a note this morning for oil market analysts, PIRA.

    Now if I believed in a constrained oil supply, I’d think such an increase in demand would be associated with a whopping big GDP number in Q4 or Q1.

    But if I were right, I’d only be lucky.

  44. Gravatar of Steven Kopits Steven Kopits
    21. November 2013 at 10:20

    Oh, and China’s oil demand is up only 2.6% yoy in the three months to October.

    That’s because China’s GDP growth is so strong now, that Caterpillar’s sales of resources equipment have been in the cellar for three quarters there. (http://www.zerohedge.com/sites/default/files/images/user5/imageroot/2013/11/Cat%20OCT%20.jpg)

    And college graduates are finding it difficult to land jobs, according to my Chinese analyst who just visited there.

    By contrast, Korean oil demand increased by 8.8% per annum during its motorization phase when GDP increased an average of 8%. So if I believed China’s GDP numbers, we should be seeing demand growth there of 8-9% or so (unconstrained supply) and 5-6% (constrained oil supply). Instead, China’s oil consumption growth is speaking to GDP growth in the 2-3% range, similar to the conclusions you might draw from the Caterpillar chart above.

    China’s oil demand growth is also lagging that of the “other non-OECD”, which is up by 3.7%. Normally, China’s growth runs ahead of the other non-OECD, but it is conspicuously lower now.

    So, maybe China’s economy is going great guns, but it doesn’t leap off the page at me.

  45. Gravatar of Mark A. Sadowski Mark A. Sadowski
    21. November 2013 at 10:41

    Steven Kopits,
    “Oil, it turns out. Demand is up 9.9% in the last four weeks compared to the same period last year, according to a note this morning for oil market analysts, PIRA.”

    According to EIA US petroleum consumption averaged 20,273.5 thousand barrels per day from October 26 through November 15. This is up 5.1% from the 18,870 thousand barrels per day it averaged October 27 through November 16 last year. Furthermore the standard deviation of year on year percent changes for four week average US petroleum consumption is about 4.7 points, meaning, as usual, this sounds like much ado about nothing.

  46. Gravatar of Lorenzo from Oz Lorenzo from Oz
    21. November 2013 at 14:23

    “Aggregate demand” being poorly labelled (and thinking of it as nominal income does help make the limitations of Say’s law clearer) just fits the pattern of unfortunate labeling in economics: “rational expectations” when what is meant is consistent expectations; “efficient markets” when what is meant is informed markets: “real x” which abstracts away from money while continuing to invoke its functions (since so many of those “real” transactions would not take place without money); “malinvestment” which implies some intrinsic quality when policy shifts can turn what was a perfectly reasonable investment into not so; and so on.

  47. Gravatar of Steven Kopits Steven Kopits
    21. November 2013 at 14:49

    Mark –

    EIA data is up 7.4% for the four weeks ending Nov. 15 compared to the same period a year ago.

    The 9.9% number is from PIRA, from a non-public report.

    But even a 5% increase is a big number. In BP’s Outlook, for example, North American oil consumption is supposed to fall by 2.6% from 2011 to 2015. Instead, US consumption is up 7.4% in a year? Wow. That’s a big number. You have to go back to 2010 to see a couple of weeks like that–and 2010 was the point coming off the trough of the recession. Before that, you have to go back to 1990.

    Will it last? I don’t know. I suspect it will, but four weeks does not a durable trend make.

    Why are you so ideological about this? Why is it a surprise that an economy is dependent on its monopoly transportation fuel? I don’t think this is a particularly deep insight. If we accept that, the rest is all about estimating the appropriate coefficients, and those are readily available in the historical record.

    For the BP data, click on the excel table: http://www.bp.com/en/global/corporate/about-bp/statistical-review-of-world-energy-2013/energy-outlook-2030.html

  48. Gravatar of Mark A. Sadowski Mark A. Sadowski
    21. November 2013 at 15:24

    Steven Kopits,
    “EIA data is up 7.4% for the four weeks ending Nov. 15 compared to the same period a year ago.”

    I’m pretty confident my arithmetic is correct (or rather Excel’s is). What I meant to say is consumption averaged 20,273.5 thousand barrels per day from October *19* through November 15 and that this is up 5.1% from the 18,870 thousand barrels per day it averaged October *20* through November 16 last year.

    “The 9.9% number is from PIRA, from a non-public report.”

    And that makes them a more relaible source of information?!? (Shades of Trading Economics.)

    “But even a 5% increase is a big number. In BP’s Outlook, for example, North American oil consumption is supposed to fall by 2.6% from 2011 to 2015. Instead, US consumption is up 7.4% in a year? Wow. That’s a big number. You have to go back to 2010 to see a couple of weeks like that-and 2010 was the point coming off the trough of the recession. Before that, you have to go back to 1990.”

    The standard deviation of such changes is 4.7 points. Thus changes of 5% or more in either direction happen fairly frequently. For example the last time the 4-week average consumption was up 5.1% or more yoy was September 20. And the last time it was *down* over 5.1% yoy was June 20.

    In other words 4-week average oil consumption is *volatile*. I practically guarantee you that it will be down by more than 5.1% yoy within a few months.

    “Why are you so ideological about this?”

    The shoe is very much on the other foot. I’m hyper-empirical. When someone makes a claim that I’m unsure about the first thing I do is look at the data. And so far nearly all of your claims are failing the smell test in a very big way.

  49. Gravatar of Mark A. Sadowski Mark A. Sadowski
    21. November 2013 at 15:41

    Steve Kopits,
    As soon as I fired that comment off I found an Excel error.

    Yes, the increase is 7.4%. (At least I now know we’re looking at the very same data.)

    The standard deviation is 3.3 points. So it is more rare than I realized.

    The previous time it was up by 7.4% or more yoy was May 2010. The last time it was down by 7.4% or more yoy was December 2011.

    On the other hand, prior to that you don’t have to go all the way back to 1990 to see such an increase. It happened in January 2001.

    But yes, I take this back, this is looking more convincing.

  50. Gravatar of Mark A. Sadowski Mark A. Sadowski
    21. November 2013 at 16:39

    Steven Kopits,
    There are seven quarters when 4-week average petroleum consumption was up by 7.4% or more yoy since November 1991. Here are the yoy RGDP changes in percent those quarters:

    1992Q2 3.2
    1993Q1 3.3
    1994Q1 3.4
    1997Q2 4.3
    1999Q4 4.9
    2001Q1 2.3
    2010Q2 2.7
    Average 3.4

    The average yoy change in RGDP from 1991Q4 to present is 2.6%. The greatest yoy change was 5.3% in 2000Q2. A yoy change of 3.4% is at roughly the 67th percentile.

    In short, quarters when a larger increase in yoy petroleum consumption occured, yoy RGDP growth ranged from just below average to excellent, and on average was better than two thirds of the observations.

    So far RGDP growth has averaged 2.13% in the first three quarters. To average 2.3% yoy we would only need RGDP growth to be 2.8% in 2013Q4, which I suppose is possible.

  51. Gravatar of Nick Rowe Nick Rowe
    21. November 2013 at 17:42

    Scott: interesting post. I decided to be contrarian, just for the hell of it, and came up with a partial defence of the “nifty new goods increase AD” argument!

    http://worthwhile.typepad.com/worthwhile_canadian_initi/2013/11/does-the-invention-of-nifty-new-goods-increase-ad.html

  52. Gravatar of ssumner ssumner
    21. November 2013 at 18:03

    Michael, Mostly irrelevant.

    Doug, The AD curve is increasing by the quantity demanded is not.

    Steven I guarantee China is growing much faster than the US. Maybe not 7.5%, but faster than us.

    Lorenzo, Good observations.

    Thanks Nick.

  53. Gravatar of ssumner ssumner
    21. November 2013 at 18:11

    Lorenzo, BTW, very good observations on conservatism. I forgot to mention that after you had a link in an earlier post.

  54. Gravatar of JohnB JohnB
    22. November 2013 at 04:24

    Scott,

    My argument is that the unemployment number means nothing when there has been zero recovery in labor force participation. The unemployment number falling simply indicates people no longer bothering to register as unemployed. I am well aware that the unemployment rate has fallen and that real GDP is up from the previous peak.

  55. Gravatar of Steven Kopits Steven Kopits
    22. November 2013 at 05:49

    Mark –

    OK, now we’re getting onto the same page.

    If you look at those RGDP increases for the quarters you mentioned, here are my comments:

    1992-1994: The oil supply was not constrained during this period, so this was probably solid GDP growth increasing demand for oil, rather than the other way around.

    1997, 1999 was also a period of strong GDP growth for the US, and we had the collapse of oil prices somewhere in there. So again, I think oil is the dependent variable.

    2001 was recovery from recession.

    2010 was also recovery from recession, but with poor GDP numbers reflecting the continued financial aspect of the crisis, I think. We were not supply-constrained at the time, as oil prices were still low due to the collapse of demand, so it looked like a normal demand recovery, albeit with disappointing GDP growth for that part of the cycle.

    The US would be “locked out” of oil consumption growth around Q4 of 2010 (around $95 Brent, if I recall correctly).

    The current surge in consumption is coming at fairly high Brent prices but pretty soft gasoline prices in the US. The US is technically adding substantial consumption above its calculated carrying capacity around $103 Brent. That suggests real strength of demand and a return of “animal spirits”, at least with respect to oil consumption. But again, four weeks does not a long-term trend make.

    In any event, if the US could post 3.4% GDP growth for Q4 or Q1, then I think we’d all breathe a sigh of relief. Personally, because the oil supply remains constrained, I would guess that growth will be higher (because the marginal utility of incremental oil consumption is higher). So I would be looking for 4%+ GDP growth for at least one of Q4 or Q1, assuming oil demand continues at the recent pace.

    You know, I do lay out much of the background of these analyses in my Princeton presentation. I introduce the demand-constrained approach, used by the IEA, BP and Citi (and almost everyone else), and illustrate how it works. Then I introduce a supply-constrained approach, which is used by us, Phibro and to a limited extent, Weatherford. I walk through the evidence pro and contra, and then take a stab at linking these models to oil industry capex (upstream spend) and US and global GDP growth.

    Admittedly, it’s a very long deck, it combines theory with application, and it’s intended to be presented, not read. But there are a number of slides in there which are just gems–really enlightening, if you don’t know much about the industry or oil markets.

    I am happy to share it.

  56. Gravatar of ssumner ssumner
    22. November 2013 at 06:00

    JohnB, I regard the unemployment rate as a better cyclical indicator than LFPR, which was also very low in the booming 1960s.

  57. Gravatar of Mark A. Sadowski Mark A. Sadowski
    22. November 2013 at 08:34

    Steven Kopits,
    “2001 was recovery from recession.”

    The recession didn’t start until March 2001 and it troughed in November 2001.

    “I introduce the demand-constrained approach, used by the IEA, BP and Citi (and almost everyone else), and illustrate how it works. Then I introduce a supply-constrained approach, which is used by us, Phibro and to a limited extent, Weatherford. I walk through the evidence pro and contra, and then take a stab at linking these models to oil industry capex (upstream spend) and US and global GDP growth.”

    I suspect the problem with all of these approaches is they fail to consider that NGDP is not determined by real factors. Supply side shocks should produce supply side effects and these are simply not detectable from the fracking boom.

    “I am happy to share it.”

    Where can I find it?

  58. Gravatar of Lorenzo from Oz Lorenzo from Oz
    25. November 2013 at 01:23

    Thanks Scott: now I have internet at home again, am catching up on my blog reading.

  59. Gravatar of Skepticlawyer » Money, prices, assets and evasions of responsibility Skepticlawyer » Money, prices, assets and evasions of responsibility
    28. November 2013 at 03:18

    […] for this is aggregate demand (for goods and services). Just as with velocity, aggregate demand is not a clear usage, but it is the accepted one, so we are stuck using with […]

Leave a Reply