Wednesday, March 20, 2013

Hayek on “The Flow of Goods and Services”

On January 27, 1981, Hayek gave a talk at the London School of Economics (LSE) called “The Flow of Goods and Services,” on the fifty year anniversary of the first of his four lectures at the LSE in 1931 on Prices and Production.

It is a curious paper which shows the fundamental flaw in Hayek’s economic ideas that he had until the end of his life.

The fundamental flaw in Hayek’s economic theory is his emphasis on prices.

For Hayek, the “coordination of economic activities [sc. in modern economies] ... is due to our relying for guidance on prices formed on competitive markets which generate the indispensable signals which tell us what to do” (Hayek 2012: 333). Even at the end of his life, Hayek was stuck in the largely fictional neoclassical view of the role of prices in market economies: the idea that all or most prices are flexible and adjust towards their equilibrium values and tend to clear markets, and signal what to produce via profit and loss (Hayek 2012: 339, 345).

Hayek was only ever dimly aware of the existence and importance of fixprice markets (although he describes them briefly at Hayek 2012: 340 then dismisses the subject completely!). Hayek never understood the fact that in very many markets changes in supply are caused by “quantity signals,” not price signals. Quantity signals are (1) changes in the amount of the stock a business carries or (2) changes in demand.

As Nicholas Kaldor argued, normally in modern capitalist economies “in actual adjustment of supply and demand, prices play only a very subordinate role, if any [sc. role]” (Kaldor 1985: 25).

By the end of his life, Hayek was willing to recognise the problems with general equilibrium theory, but never abandoned his misguided and erroneous view of the role of flexible market-clearing prices:
“It is tempting to describe as an ‘equilibrium’ as ideal state of affairs in which the intentions of all participants precisely match and each will find a partner willing to enter into the intended transaction. But because for all capitalist production there must exist a considerable interval of time between the beginning of a process and its various later stages, the achievement of an equilibrium is strictly impossible. Indeed, in a literal sense, a stream can never be in equilibrium, because it is disequilibrium which keeps it flowing and determining its directions. Even an apparent momentary state of balance in which everybody succeeds in selling or buying what he intended, may be inherently unrepeatable, irrespective of any change in the external data, because some of the constituents of the stream will be results of past conditions which have changed long ago.” (Hayek 2012: 338–339).
So while a general equilibrium state can never exist in the real world, nevertheless Hayek is still thinking of some market tendency towards equilibrium.

But it is clear that flexible, market-clearing prices and wages still have a fundamental role in Hayek’s theory:
“The price ‘gradients’, as I have called them, which keep the stream moving, are signals indicating momentary and passing conditions largely determined by events of the past; and how many of the potential productive forces the stream will be able to absorb will depend both on whether enough of these price signals or constellations of signals, stand at clear, indicating that in certain directions prices of output exceed prices of input, and whether the whole system of signals favours an increase or a decrease of the volume of the whole stream, and not merely the rate of flow at its mouth. Every price which is slow in adjusting itself to changes in local conditions may choke a particular streamlet and thereby impede the continuous movement of the whole. The degree to which the order will ever approach the unattainable ideal of equilibrium will depend on the speed of adaptation and of the communication process which brings it about.

This is much more important than the degree of perfection towards which it may tend: since the process of adaptation never ends, the average closeness to the ideal will be determined by the speed of the response to changes in the data. The fluctuations which are inevitable in an order brought about by feedback processes, which social structures share with biological ones, will always cause divergences from stability and it will be the speed, not the ultimate perfectness of adaptation which will determine the steadiness of the flow. Prompt movements in the right direction will do more to keep the stream flowing than any precise calculations of equilibrium conditions could do whose results would probably come too late to improve the flow. Inflexibility of prices, particularly of the prices of labour, can block many of the arteries on which the nourishment of society, and especially of the very people who endeavour to maintain their relative position, depend. Unemployment is not so much a function of ‘aggregate demand’ as of the elasticity of the price structure which, of course, all striving for a ‘just’ distribution obstructs.” (Hayek 2012: 339).
So the “order” in a capitalist system is still dependent on flexible prices and wages, and Hayek is here thinking of a tendency towards a market-clearing wage and price vector. Prices are an informational system that allow movement to greater coordination in a market system. Rising prices signal that more needs to be produced of a particular commodity. Falling prices that less needs to be produced. Price adjustment towards a stable level indicates supply and demand is coming into equilibrium.

Yet the reality of modern capitalist production is widespread fixprice, not flexprice, markets (though obviously flexprice markets do exist). The whole Hayekian system is built on a house of sand in terms of its price theory.

It is sometimes asserted that Hayek rejected general equilibrium theory towards the end of his life (Caldwell 2004: 226-227).

The closest thing to a repudiation appears in Hayek’s statement in an interview as transcribed in the book Nobel Prize-Winning Economist: Friedrich A. von Hayek (1983, pp. 187-188):
“HIGH: To what extent do you think that general-equilibrium analysis has contributed to the belief that national economic planning is possible?

HAYEK: It certainly has. To what extent is very difficult to say. Of the direct significance of equilibrium analysis to the explanation of the events we observe, I never had any doubt, I thought it was a very useful concept to explain a type of order towards which the process of economics tends without ever reaching it. I’m now trying to formulate some concept of economics as a stream instead of an equilibrating force, as we ought, quite literally, to think in terms of the factors that determine the movement of the flow of water in a very irregular bed.”
Yet it is difficult to really find a clear and unambiguous repudiation of general equilibrium theory, because Hayek still adhered to the equilibrating role of market-clearing prices in his economic analysis.

What Hayek did believe – in contrast to Walrasian general equilibrium theorists – is that mathematics and Walrasian theory could not be used “determine and predict” the numerical values of a general equilibrium state beforehand by means of, for example, supply and demand equations (Lachmann 1983: 374). This was already stated in the The Pure Theory of Capital (1941), where Hayek asserted that it was necessary to “abandon every pretence that [sc. equilibrium] … possesses reality, in the sense that we can state the conditions under which a particular state of equilibrium would come about” (Hayek 1976 [1941]: 28).

Yet even in a world of disequilibrium prices, Hayek still believes that entrepreneurial activity drives these prices at least towards their market clearing values, even if one cannot predict such values with the Walrasian simultaneous equations of general equilibrium.

The same basic idea informs this statement Hayek made on April 9, 1975 in a talk to the American Enterprise Institute in Washington DC:
“The primary cause of the appearance of extensive unemployment, however, is a deviation of the actual structure of prices and wages from its equilibrium structure. Remember, please: that is the crucial concept. The point I want to make is that this equilibrium structure of prices is something which we cannot know beforehand because the only way to discover it is to give the market free play; by definition, therefore, the divergence of actual prices from the equilibrium structure is something that can never be statistically measured.” (Hayek 1975: 6–7).
BIBLIOGRAPHY

Caldwell, B. 2004. Hayek's Challenge: An Intellectual Biography of F.A. Hayek University of Chicago Press, Chicago and London.

Hayek, F. A. von. 1976 [1941]. The Pure Theory of Capital. Routledge and Kegan Paul, London.

Hayek, F. A. von. 2012. “The Flow of Goods and Services,” in H. Klausinger (ed.), The Collected Works of F. A. Hayek. Volume 8. Business Cycles. Part II. University of Chicago Press, Chicago and London. 331–346.

Kaldor, Nicholas. 1985. Economics Without Equilibrium. M.E. Sharpe, Armonk, N.Y.

Lachmann, L. M. 1983. “John Maynard Keynes: A View from an Austrian Window,” South African Journal of Economics 51.3: 368–379.

23 comments:

  1. "Yet the reality of modern capitalist production is widespread fixprice, not flexprice, markets (though obviously flexprice markets do exist). The whole Hayekian system is built on a house of sand in terms of its price theory."

    Even if every capitalist based his prices on costs + mark-up and every adjustment was eventually predominately in quantities and not in price then analysis shows that economic adjustments can be explained by a Hayekian model but not by a post-Keynsian one.

    Take 2 goods (A and B) where the price of each is exactly equal to the cost of production plus a markup. Each good consists of labor plus a capital good combined in different ways.

    Case 1: Suppose a supply shock causes the supply of the capital good to be reduced and the price to increase. How will this reduced supply be distributed between good A and good B? Hayekian theory shows the process by which pricing signals in the consumer markets will be used to allocate the changed quantity of the capital goods between good A and good B (the transfer of the additional costs of inputs onto the costs of output.) The new equilibrium prices will still be cost+markup but entrepreneurial activity and pricing signals will be tools used to achieve this.

    Case 2: A change in demand for good A relative to good B. This will result in a different distribution of the capital good between A and B. The ability to sell more of good A will lead the producers of this good (even if they use cost+markup) to bid away the capital good from producers of good B and lead to explainable changes in the price and qty of both goods.

    In both of the above cases the end result will be cost+markup and the adjustment will likely be in quantity rather than price - but a Hayekian process explains how this result is derived.

    So: For Post-Keynsians who claim that Hayekian theory is built on sand I have a question. Where is their theory that explains how prices and quantities produced are derived beyond the mere assertion that "it is all down to administered prices" ?

    ReplyDelete
    Replies
    1. (1) Case 1:
      "Suppose a supply shock causes the supply of the capital good to be reduced and the price to increase. "

      It does not follow the price will increase. Reduced supply may well mean just bottlenecks, delays in delivery and unavailability of that good.

      You are just begging the question: assuming the truth of flexprices in all markets.

      Case 2:
      "The ability to sell more of good A will lead the producers of this good (even if they use cost+markup) to bid away the capital good from producers of good B and lead to explainable changes in the price and qty of both goods.

      It does not follow that prices will change. Quantity adjustments in both good A and good B can occur without necessary price changes.

      (2) finally I never said "it is all down to administered prices". Obviously in flexprice markets you have a more conventional story being played out. Though even here the story is not quite the same as in neoclassical theory (see Kaldor, 1985. Economics Without Equilibrium. M.E. Sharpe, Armonk, N.Y., pp. 14-23).

      But administered prices are a big part of the picture and Hayekian theory ignores them.

      Delete
  2. The logic of my 2 cases would not be affected if prices did not change at all (and in fact this is quite likely in the second case) - so your criticism on that point is irrelevant.

    However as relative prices clearly do change in the real world any worthwhile economic theory needs to be able to account for both qty and price changes - and this surely has to factor in changing costs and/or demand - which is what the Austrian theory is able to do

    I'm not sure I fully understand what is meant by administered price. Do you mean prices are set by an external agency like happens in a few industries, or are administered by businesses themselves based on some sort of pricing formula ? If is the former then obviously that's just a market intervention which will clearly hamper any Hayekian processes. If it is the latter then one still needs a theory to explain the costs that the pricing formula is based upon and the level of output that it results in. As I tried to demonstrate EVEN IF ONE ASSUMES LARGELY ADMINISTERED PRICES then the Hayekian framework explains costs and quantities in a way that nothing I have seen in the PK framework is able to.

    I struggle to believe though that Apple, Walmart, the various sellers who compete on Amazon, and even car manufacturers are using administered pricing.

    ReplyDelete
    Replies
    1. So you are ignorant of what "administered prices" even means, yet are confident that "EVEN IF ONE ASSUMES LARGELY ADMINISTERED PRICES then the Hayekian framework explains costs and quantities in a way that nothing I have seen in the PK framework is able to."!?

      http://socialdemocracy21stcentury.blogspot.com/2013/03/kaldor-on-economics-without-equilibrium.html

      http://socialdemocracy21stcentury.blogspot.com/2012/06/price-rigidity-in-new-keynesianism-and.html

      http://socialdemocracy21stcentury.blogspot.com/2012/07/more-on-prices-in-real-world.html

      http://socialdemocracy21stcentury.blogspot.com/2012/08/lachmann-and-post-keynesianism-on-prices.html

      http://socialdemocracy21stcentury.blogspot.com/2012/12/mises-versus-lachmann-on-equilibrium.html

      Delete
  3. I looked at the Kaldor model and it seems flawed.


    Kaldor identifies some quantity signals

    (1) a change in the amount of the stock/inventory* a business carries,
    (2) a change in demand: changes in the sales volume or orders, or
    (3) a combination of (1) and (2). (Kaldor 1985: 23).

    Lets assume that is what happens in my simple example for good A.

    The producer of good A increases his demand for inputs and the supplier of those inputs sees similar quantity signals and increases production, which sends quantity signals to his suppliers an so on backwards.

    Now this is great as long as long as all supply chains are perfectly scalable - but the main premise in economics is that they are not - we live in a world of scarcity , and these scarce resources need to be allocated somehow.

    What happens when a quantity signal runs up against a resource constraint? In my model suppose the entire supply of the capital good is being used up before the demand for good A increases? Kaldor's model breaks at this point. You can't expand qty for A without decreasing qty for B.

    I can think of only 2 ways to resolve this. 1) Use of the price system to allocate scarce goods or 2) rationing by a central authority.

    What am I missing ?



    ReplyDelete
    Replies
    1. "What happens when a quantity signal runs up against a resource constraint? "

      In a real world capitalist economy? Really excessive demand might cause bottlenecks, delays in delivery and unavailability of certain goods.

      I have seen myself this in supermarkets: a natural disaster reduces supply of some good. The price does not soar.

      What you see is a sign: "this good is temporarily unavailable due to supply issues" or something of that sort.

      Now when you get to flexprice markets for certain factor inputs, certainly that is when demand may well cause price changes, and your Hayekian story might be relevant.

      But even here as I have said above, the story is just isn't the same as in neoclassical or Austrian theory (see Kaldor, 1985. Economics Without Equilibrium. M.E. Sharpe, Armonk, N.Y., pp. 14-23), and throughout the fixprice markets Hayekian theory does not work.

      Delete
    2. So in a fix price model if demand increases but supply can't because of a resource constraint then the result will just be a shortage and no mechanism exists to adjust the price of either the good or its inputs ?

      That's seriously part of post-Keynsian price theory ? .

      Delete
    3. I have already said above that in flexprice markets prices might increase.

      Delete

    4. I can see how Kaldor's idea of quantity signals is a useful addition to Hayekian price signals but can't see how it can eliminate the need for price changes.

      Can you give me some examples of goods whose markets you would define as fixprice ?

      Delete
    5. So do you have any examples of fix price markets ?

      Delete
    6. Fixprice markets exist in most manufacturing, many service and retail industries.

      Any market that is not organised like an auction with bids, or a bazaar where you haggle is likely to be a fixprice market.

      Tell me: do you find yourself bidding or haggling for goods in department stores or supermarkets? Or do you find fixed prices?

      Even with the exception of clearance sales, fixed prices are the norm.

      Delete
    7. So fix price just means fixed at the point of sale ?

      I'm pretty sure that most economists (including Austrian ones) are aware that this is how prices work in the real world. This is totally consistent with Hayekian pricing theory - its amazing that you could think otherwise.

      The issue is not whether prices change day to day with every fluctuation in demand or supply conditions - but rather whether there is a tendency in the market for these changes to ultimately lead to equilibriating prices and quantity changes. (BTW: Austrian theory explains how if the change is on the demand side then often the result will be a quantity and not a price change).

      Delete
    8. "So fix price just means fixed at the point of sale ?"

      No, it does not mean just that.

      It means fundamentally: prices are administered or "set" by corporations and businesses before sales, according to normal production costs plus a profit markup.

      Businesses leave their prices unchanged for significant periods of time, from three months to a year, despite changes in demand When prices are adjusted this is the result of changes in factor input costs, including raw materials and labour. Changes in the profit markup result from competition and need for profit.

      The full post here:

      http://socialdemocracy21stcentury.blogspot.com/2012/07/more-on-prices-in-real-world.html

      Delete
    9. I looked at that earlier post.

      I think that your 4 points are empirically true in the real world to varying degrees. I don't really see this as being a huge challenge to Hayekian pricing theory though which is about tendencies towards equilibrium not day-to-day actual equilibrium. Such tendencies can clearly exist in a world of short-term "fix price" (where all of your 4 constraints exists).

      In fact as Hayekian theory is largely focused on what can go wrong with the pricing system (ABCT) as well as what can go right I think an understanding of these fixprice models would add and not detract from the Austrian framework.



      Delete
  4. of course quantity signals play a large role. Yet even quantity signals, if you break them down, are composed of smaller units of goods that have there own distinct prices.

    Of course there are "fix prices" , "sticky price"s in the NK parlance. Yet in the long run, if the ask price of a seller does not match the desired bid price of the buyer, "quantity signals" will show large surpluses and inventories that lose money for most businesses. So in the long run, consumers rule.

    Your attempts to deny the law of demand are stretching, to say the least. Even in the case of a company that has IR, its cheaper for that company to raise the price of its products within the range of possible prices that a buyer is willing to pay, than to increase production. So your attempt fails. (Unless its at its upper bound, and any further increases in prices will decrease demand. But that in itself, is no violation. The law of demand has a side for sellers and for buyers, and sellers have to keep both in mind to maximize profit.}

    For me, equilibrium is about two things, Firstly, about turnover and transactions, secondly, about the tendency for profits to disappear, and setting things to approach their cost of production.

    In the first case, we see equilibrium around us every day. Not only is there a "tendency" to approach it, it happens! Every measure of revenue and sales pre assumes the fact that buyers and sellers reached an agreement. [By the way, I think there are multiple equilibria, I don't think the simple neoclassical story is accurate) the second thing, is manifestly untrue.

    ReplyDelete
  5. Re-posted from previous post-



    "Furthermore, how could 19th century governments expand base money when they were forced to use scarce gold? It is precisely fiat money that is required to allow this type of intervention to work smoothly and reliably."

    {Sigh.} I provided you with a mechanism. Given that the gold standard era was a mix a of fiat money and gold, and given that governments had the power to define the smallest unit of commodity money, a good way to expand money would be to raise the price of gold/silver in a financial panic in terms of credit money. For example, suppose an ounce of gold was worth ten pounds in England, and five in America. A financial panic would see the Treasury departments of both countries RAISE the gold price, thus creating more money in effect. Back then one had to unfortunately work with the "mystique" of the gold standard" My solution would have been the cheapest most effective way to maintain demand back then.

    "(1) If by NGDP targeting you mean expand base money by the % you want GDP to grow by, this is just other monetarist "pushing on a string" solution.

    You might stave off financial collapse that way, but if expectations are shocked and demand for credit collapses, it is virtually useless for expanding private investment."


    Really? Do you deny that central banks have the power to pay any price they wish for any assets they wish? i don't know the rules in the UK, but in the U.S. the Fed has the power to buy anything it wants, at whatever bid it wants. If it wants to set market bond yields below zero, it can. It can pay 12 trillion for ten trillion dollars worth of mortgages, thus setting interests rates at -20% Banks COULD PAY borrowers a fee, say 4 percent, for the privilege of accepting a loan from them! Those banks could sell the securitized loan to the Fed and still make a profit.
    It is truly amazing to read Keynesians skeptical of monetary policy. You have to deny all sorts of obvious truths, and perform mind boggling mental contortions in your mind. If your position was true, LK, then The Fed could buy all the U.S. gov debt IN THE WORLD, all 13 trillion, with no effect on aggregate demand, and even if it were so, and it would still be something worth doing because it would be an amazing free lunch. (The same goes for private debt.)

    Oh, and on NGDP targeting, read Scott Sumner, Nick Rowe, lars Svensson, people much smarter than I am, to explain things. I disagree with some of the things they have to say, for example, I think central banks have to buy tens of trillions to expand NGDP, or it at least credibly threaten to do so to change expectations. whereas they think all a CB has to do is announce a change in targets. But the focus on NGDP which is a measure of MV or PY. or just plain aggregate spending, is the right path to take. Todays monetarists AGREE with Keynes, that spending and investment drives the economy, not just plain M1, M2, or M3

    ReplyDelete
    Replies
    1. If monetary policy had the influence you think it does, then QE1, QE2 and QE3 should have had a very big influence on private investment and employment.

      "Oh, and on NGDP targeting, read Scott Sumner, Nick Rowe, lars Svensson, people much smarter than I am, to explain things. I disagree with some of the things they have to say, for example, I think central banks have to buy tens of trillions to expand NGDP"

      Tens of trillions? So in other words you admit that Scott Sumner, Nick Rowe's etc NGDP targeting ideas have already been refuted in the real world?

      Delete
    2. Jan: Well i don´t think it´s fair as Edward do, to lump together Lars.E.O Svensson with Scott Sumner etc ,since he is an advocate of "negative interest rate" and fiscal and a monetary dove,and been advocate for a more active especially monetary policy to bring down unemployment as deputy governor of Swedish
      Central Bank.He been in opposition to governor Stefan Ingves more restrictive policy.I don´t think one could call Svensson a "monetarist" new or old,style,he is has his own ideas. https://larspsyll.wordpress.com/2013/02/06/on-inflation-targeting-and-rational-expectations/

      Delete
  6. "...all or most prices are flexible and adjust towards their equilibrium values and tend to clear markets, and signal what to produce via profit and loss"

    Even if they did, that doesn't mean these processes would lead to optimum outcomes.

    ReplyDelete
  7. "If monetary policy had the influence you think it does, then QE1, QE2 and QE3 should have had a very big influence on private investment and employment.

    "Oh, and on NGDP targeting, read Scott Sumner, Nick Rowe, lars Svensson, people much smarter than I am, to explain things. I disagree with some of the things they have to say, for example, I think central banks have to buy tens of trillions to expand NGDP"

    Tens of trillions? So in other words you admit that Scott Sumner, Nick Rowe's etc NGDP targeting ideas have already been refuted in the real world?"

    They are partly wrong of course. But that doesn't refute the whole of their ideas. Just like mainstream Keynesians failing to anticipate stagflation doesn't repudiate keynes in its entirety. And QE1, QE2, QE3, is like trying to fight a firestorm with a squirt gun. The stimulus was to small! (Lol, i have something in agreement with Keynesians like PK here, except about MP rather than FP)

    ReplyDelete
    Replies
    1. Fine. You think QE needs to be tens of trillions.

      Given the US economy is about $15 trillion, would you say $20 trillion or $30 trillion is the appropriate level of QE?

      Why not just spend 2-3 trillion in the form of stimulus over a few years on badly need public infrastructure?

      Delete
  8. A seller can fix his price all he wants, doesn't mean he'll stay in business, doesn't mean he won't make a profit. The price is always "flexible" but entrepreneurs don't always have to change theirs.

    ReplyDelete
  9. How are changes in quantity and demand reflected in a market? Prices!

    ReplyDelete