As I Begin to Understand Macroeconomics...

I begin to see how Say's Law is perhaps the most important concept in the subject. In saying that, I am not contending that Say's Law* always holds (nor am I claiming it doesn't!): I am suggesting that many (most?) macroeconomic disputes can be understood as disputes over its applicability.

If we pay attention to the Keynesian model, we see that the "paradox of thrift" occurs when S (savings) > II (intended investment). The model contains no hysteresis, so it makes no difference how that condition came about: it could be a increase in savings not matched by an increase in intended investment, or it could be a drop in intended investment not matched by a drop in savings.

The "right" way to deny that this paradox is of importance is not to say it implies savings is bad, because it doesn't, but to try to show that II shifts along with S, which Bob Murphy understands: "In contrast, a Rothbardian (say) is going to argue that when people save more, investment will go up, and so we’re not in trouble."

Yes, he's got it here! A gap between S and II would generate this "paradox," but per Rothbard (and all advocates of the classical understanding of the market for loanable funds), such a gap won't arise, because a change in S will also change II.

Now, we've gotten to the real issue under contention: do changes in S and changes in II (generally) endogenously equilibrate, or don't they? If they do, Say's Law holds, and the classical view is correct. If they don't, Keynes is correct, and the market economy is not inherently stable.

* As Thomas Sowell notes in his great book on the history of Say's Law, there are over half-a-dozen different things this law might be taken to claim.


  1. Anonymous7:51 PM

    I personally don't think the term "intended investment" is helpful, because it doesn't really exist. There are a lot of things that I might intend to do, but unless I do them, they didn't happen. I must ask, whose intended investment are you speaking of, and by what measure are we to determine that intended investment differs from actual investment?

    Unfortunately, I am a drained at the moment due to debating what constitutes savings, so I'll probably have to continue these thoughts on investment tomorrow. Well, unless this glass of scotch rekindles my fire. ;)

    1. "I personally don't think the term "intended investment" is helpful, because it doesn't really exist."

      Joe, please study the model: you'll see this term has a perfectly clear meaning in the model.

    2. You're saying that intended investment doesn't have any independent reality - it's just something our brains come up with to navigate the world around us?

    3. Anonymous8:36 PM

      You're right. I probably should get a better grasp on the model itself before I go spouting off, it's been a while since I've actually studied it.

      My instinct tells me that this mismatch has something to do with how Keynes defined saving (well, *one* of his definitions, anyway). Certainly, I could be wrong, but that is just my first instinct.

  2. Close. Understanding Say's Law/Walras' Law in a monetary exchange economy, when it/they do and do not hold, is indeed the most important question in macro.

    May I offer my thought on this subject?

    1. I have to disagree. The most important question in macro economics is, When will people start or stop making future promises?

      Keynes released that government spending could be substituted for promises. We are not testing, for how long?

      As a growing number of people stop making future promises, the rate interest begins loosing its effectiveness, until a point is reached where the interest rate has no effective whatsoever. As Coase just wisely explained in the HBR the answer to the stop/start question lies in the "influences of [psychology], society, history, culture, and politics on the working of the economy," which economics has been ignoring.

      In the 1930s people had to stop making future promises because they had the wisdom to see where the Nazis, Japanese, and Stalinists were going. That reality never appears in any economic model of the Great Depression.

    2. Interesting post, Nick. I am digesting it.

      I think as Malthus and Sismondi understood a general glut, it is possible without money. I explain how here. (You may think "BUt that is not why I / modern macroeconomists mean by a general glut!" I would be happy to concede that point. I am just trying to grasp what M & S meant.)

  3. This is an excellent point. I found this interesting: "Yes, he's got it here! A gap between S and II would generate this "paradox," but per Rothbard (and all advocates of the classical understanding of the market for loanable funds), such a gap won't arise, because a change in S will also change II."

    This is actually much like the Keynesian story which also assumes the gap won't arise in actuality. The identity has to hold (hence talking about intended I). But different assumptions are made about what will adjust. Keynes makes no commitment either way. Some people hold that we are guaranteed to have only pleasant increases in I to ensure that the identities hold.

  4. Say's Law/Walras Law is indeed the most important topic in macro. SL/WL is, in general, wrong. But Keynes was also wrong about why it was wrong.

    An excess of desired saving over desired investment will *not* create a general glut, unless it takes the form of an excess of desired saving in the form of medium of exchange.

    Suppose everybody wants to buy land, (which is desired "saving") and land prices do not adjust to clear the market in land. So what. They can't buy land, because nobody wants to sell. So they have to buy something else instead. No general glut. Unless they decide they want to hold money instead.

    1. re: "An excess of desired saving over desired investment will *not* create a general glut, unless it takes the form of an excess of desired saving in the form of medium of exchange."

      Well right. Isn't that why Keynes so painstakingly charts out liquidity preference?

    2. Nick,

      This is precisely how I understand Say's law. It assumes monetary equilibrium in order to function as increased supply creating demand for other goods and thus opening a "vent" to new markets.

      But I think it's value is deeper, as evidenced by the utter bedlam unleashed when Say's law is disregarded (ie, Keynesian claims that "consumption increases output"). All demand which isn't the result of inflation is the result of effective production. As the old-school classicals said commodities exchange with commodities. There is no autonomous demand or consumption that isn't enabled by production.

      Once you understand this, the entire notion of consumptionist solutions to recession and claims that consumption can grow the economy are instantly recognizable as a fallacy. The economy grows when we effectively get better a producing value. Consumption is the ends, not the means. It's never the means.

      Cash for Clunkers made society poorer. So does UI benefits. They aren't "stimulus". They may be the right thing to do morally, but we're poorer so long as someone is consuming and not producing.

      If people are increasing their holdings of exchange media, our monetary system should meet that demand such that nominal spending is stable. Doing so prevents bad stuff but doesn't ensure good stuff, like full employment or growth. Say's law is a deep insight about where our ability to consume and grow is actually derived.

    3. John, wait a second, you say, "This is precisely how I understand Say's law," when Nick says "SL/WL is, in general, wrong."

      But you seem to want to claim it is always correct!

    4. When have you seen me write that savings always equals investment, Gene? I subscribe to a monetary equilibrium approach and have since 2009 when I really started reading into this stuff. I put Larry White talking about Hayek's policy norm as stabilizing "MV" in one of our videos.

      Savings equals investment provided that the savings are not being hoarded in stockpiles of exchange media nor trapped in a broken banking system (thanks to a combination of paying interest to sit on reserves, regulatory capital requirements, etc).

      Looking at NGDP, though flawed, seems to be a pretty reasonable gage of the stance of monetary policy and monetary equilibrium.

      So all of that is the asterisks against Say's law insofar as production opens a vent to other markets. But even with those astericks, the fact remains that so-called "demand" can only be separated by "supply" in a nominal sense. Economic transactions are exchanges between two producers. As the classicals put it, goods buy goods, commodities exchange with commodities. Time and money can mask this fact but they don't eliminate it. The only time there can truly be "demand" without there first being successful value-adding supply is through printing new money.

      So I get it. Money is a loose joint. We should, from a policy standpoint, aim to stabilize that joint through policies or institutional arrangements that promote nominal spending stability. But there is no reason whatsoever to bias our institutions and policies such that they encourage consumption and discourage savings.

    5. "When have you seen me write that savings always equals investment, Gene?"

      Now I am guessing from the last comment that this is John Papola commenting. Well:

      1) How was I to know this from "John"? It is not an uncommon name!
      2) Frankly, even if I did, I really only know your views from your videos.

      "The only time there can truly be "demand" without there first being successful value-adding supply is through printing new money."

      That "first" in their shows you don't actually get Say's Law!

    6. 1. That is a remarkable guess! It's Papola. For some reason "blogger", though owned by Google, didn't pull my full name. I've fixed that now.

      I wasn't attempting to presume that you follow my every move. I was just saying that your suggestion that I always think Say's law is correct isn't right and I didn't say that it was.

      2. I suppose I don't really understand Say's law in the way that Nick Rowe espouses it (as "Walras' Law") if the caveats above aren't correct. So could you please help explain where you believe I'm getting it wrong?

      My understand of Say's law is as follows:

      Demand is constituted by the supply of goods which can be sold at cost-covering prices. And since goods buy goods with money as a medium of exchange, the "sold" part isn't a chicken-or-the-egg demand is baked in the definition of where demand comes from, but rather an acknowledgement that both sides of the trade are coming to the table having produced goods and services which can be traded with one another.

      When is demand not, ultimately, first constituted by supply, Gene?

    7. "When is demand not, ultimately, first constituted by supply, Gene?"

      OK, here is where I think this analysis goes wrong: supply is not "first," because my supply of, say, teaching, IS my demand for the goods I can acquire from the pay I get for teaching. Supply and demand are the SAME thing seen from two different perspectives, not two things where one can come before the other.

      To DIFFERENTIATE, notionally, supply and demand is analytically useful. But to DIVIDE them and think that one can occur without the other is like thinking we could have a coin with a head but no tail.

    8. So Gene, you teach WHILE buying goods and services? Or do you teach, get a paycheck, a THEN go out and demand goods and services?

      Production ("supply") does come first in a very real, practical sense.

      What I see at the truly important lesson to take away from Say's Law is that unproductive consumption of resources is never the path to growth (hence my video). Cash for Clunkers made society poorer. Growth comes from increasing productivity. The aggregate spending approach masks that dynamic and so we end up with a so-called "economics" where we can consume our way to plenty and alien invasions are a stimulus.

      That lesson from Say's law withstands monetary disequilibirium. Excess demand for the medium of exchange? Increase the supply. Don't push people to consume more nor redistribute income towards those with a higher "marginal propensity to consume".

      But, in a deeper sense, I totally agree with you that, at least in the aggregate, supply and demand are the same thing. There are no aggregate supply and demand "curves". There is only the total volume of exchange and at the point of exchange, we have two suppliers who are at the same time two demanders.

    9. "So Gene, you teach WHILE buying goods and services? Or do you teach, get a paycheck, a THEN go out and demand goods and services?"

      So John, you still have a long way to go, hey? Well, Rome wasn't built in a day.

    10. I was honestly hoping to learn from your replies, Gene. Why the snark? Insulting people is an odd educational approach for a professor, though I have seen it. Sarcasm tends to work better and is generally less obnoxious in person when you can look the person in the eye.

      I'd still like to learn why you think my explanation is incorrect if you're willing to try. Less snark would make it more pleasant.

    11. "Why the snark?"

      Because my previous comment fully handled your second-to-last one, and yet you ignored it and posted as if I hadn't written it, "pointing out" to me that I work before I get a paycheck and go shopping. Treat me like an idiot, you get snark. And then, apparently, cry like a five-year old.

      "Insulting people is an odd educational approach for a professor..."

      You know what: I ain't your professor.

  5. Since we know that markets are not inherently stable, we know that Say's law is wrong.

    That's science, straight from Richard Feynman.

    1. Anonymous8:16 AM

      I can't tell if you're being sarcastic or not. But in case you aren't, we know that markets are not inherently stable? Austrians would dispute this and point to government intervention in the past that many economists don't account for. Ironically, Feynman always believed that professionals in the social sciences didn't do the work necessary to prove what they believed they knew. Your statement would certainly be a perfect example of that.

      At least the type of inherent stability needed to prove Say's Law. Nothing in the economy is literally 'stable'. After all, human action itself is change.


    News-item from Nov 2011:

    Corporations have a higher share of cash on their balance sheets than at any time in nearly half a century, as businesses build up buffers rather than invest in new plants or hiring.

    Nonfinancial companies held more than $2 trillion in cash and other liquid assets at the end of June 2011.......Cash accounted for 7.1% of all company assets, everything from buildings to bonds, the highest level since 1963.

    Considering that the economy is not at its maximum productive capacity, why is this cash accumulating? BTW, companies were accumulating cash well before the fiscal crisis. Does Say's Law have anything to say about it? To quote Wiki, "In Say's view, a rational businessman will never hoard money; he will promptly spend any money he gets "for the value of money is also perishable." Really?

  7. "and all advocates of the classical understanding of the market for loanable funds"

    Wait, why can't one believe that interest rates equilibrate savings and borrowing AND believe that hoarding cash under your mattress is deflationary and won't contribute to the supply of loanable funds? Or am I misunderstanding this claim, Gene?

    1. If you believe the latter you are saying liquidity preference AND the interest rate equilibrate savings and borrowing.

  8. What about the view, which I might hold, that savings is (non-trivially) a species of investment, and so the whole distinction (and thus the possibility that S =/= II) is spurious?

    That is, someone who saves is investing in an asset that yields option value and insurance against uncertainty. They are also (in expectation) decreasing the "consumption load" on the economy's real resources, just like "real" investments.

    I've noticed that most of the Keynesian and market monetarist positions only work by assuming that "uninvested" savings (i.e. savings not invested by buying things from others) are a pure drain on the economy, and thus any conversion of it to something else is a gain. (at least in "underspending"/"idle resources" situations)

    Such a position would also be forced to distinguish between investment in others and investment in one's self. That is, if I hire a worker, out of my savings, to research an idea, that's "investment", and thus helps S equal II. But if I "invest" the same money in a reserve fund for myself, so that I don't have to work at a job while I research the idea myself, that's "not investment".

    Taken seriously, this would mean than *any* holding of cash between when it's received and when it's spent, in which the holder doesn't know what they'll spend it on, is economically harmful and would ideally not exist. But once you accept this, you must deny one of the functions of money, and the role of uncertainty in giving it value. You would have to hold up Weimar as the ideal, where people spend the moment they earn.

    1. "What about the view, which I might hold, that savings is (non-trivially) a species of investment, and so the whole distinction (and thus the possibility that S =/= II) is spurious?"

      No, Silas, look at the model: you have just given an excellent explanation of why S = I. II is different!

    2. Silas, sorry you're comment disappeared. You have to look at what II means in the model.

  9. Sorry, I looked at the model, and I don't see it. I do appreciate the confirmation of my suspicions though.

    1. You did not look that carefully, then, because II in the model is DEFINED so as to exclude hoarding. I'm not saying your different definition is wrong, but that you don't seem to realize you are operating on a different definition than the model is.

  10. I think the key to Keynes' critique of Say's Law is liquidity preference: The nominal interest rate must equalize money supply and demand, thereby preventing people from hoarding. So imagine the loanable funds market with Saving and Investment as functions of the real interest rate. Draw saving when Y equals potential output. The nominal rate comes from money demand and supply. Subtracting inflationary expectations, suppose we get a real rate that corresponds with the rate that equalizes full employment saving and investment. Now suppose investment slumps. In a classical model, r falls to maintain the equality of S and I. Period. For Keynes, r can't fall unless i falls, and i doesn't fall without an excess supply of money, so we have S>I right now. With flexible prices, S>I would mean falling prices - since there is an excess supply of goods and services. This fall in P will increase the real money supply (M/P), pushing the nominal and real interest rates down until we reach the new natural rate. So with flexible prices you can get the Classical result. But note that even here, the mechanism is non-Classical: the classical dichotomy is broken: adjusting to the slump in Investment requires falling prices. You can't just point to the loanable funds diagram and say, "See, r will fall." After Keynes, even opponents of Keynes adopted this non-classical view of the mechanism that will, in the best case, keep S=I. And of course Keynes thought that the adjustment comes with a fall in income, and thus a leftward shift of saving, not with falling prices.

    1. Kevin, that is how I am grasping things at this point too.

  11. Jebus guys. Say's Law is not that hard. But talking about it will drive you nuts. You just need to do the math.

    Y=C+I (Product, e.g. GDP)
    Y=C+S (Income, e.g. GDI)

    Subtract C=>

    I = UI+PI (Unplanned Investment + Planned Investment)
    S = UI+PI

    and at equilibrium:
    UI = 0
    S = PI+UI=PI+0=PI
    => S=PI(at equilibrium)
    When UI!=0, it is trivially true that S!=PI, because
    PI 0)
    PI>S(if UI < 0)

    Unplanned investment is basically inventory overshoot and undershoot.(yes it can be positive or negative) I don't think that anyone will have any problem with inventory overshoot. (A company purchases X widgets for the next quarter. Only Y widgets are purchased with Y < X.)

    Inventory undershoot is a little harder to understand. How can a company sell more widgets than it produces? It can convert investment goods into consumption goods.(For example, by overrunning its production lines. The factory will need to be replaced sooner, but more widgets can be produced.)

    That should handle Say's Law and our accounting identities.

    Now the rub, modern Keynesian theory only requires that PI!=S when the economy is at disequilibrium. Now note that this model is static. It doesn't allow Y to ever change.(but clearly, Y does change...) Derive how it is possible for the equilibrium to shift and you've just derived the income multiplier, ie "multiplier effect". Hint: Planned investment is based on past consumption. Savings is based on past income and the marginal propensity to consume.

    Dispute that if you want, but please, please get the theory right.

    1. Unknown, OK, that all looks sensible, but who is "getting the theory wrong," in your view?


Post a Comment

Popular posts from this blog

Central Planning Works!

Fair's fair!

More college diversity and tolerance