Home > Uncategorized > What is ‘effective demand’?

What is ‘effective demand’?

from Lars Syll

Samfundsøkonomi er en svær disciplin – også for topøkonomer - FINANSEconomists of all shades have generally misunderstood the theoretical structure of Keynes’s The General Theory. Quite often this is a result of misunderstanding the concept of ‘effective demand’ — one of the key theoretical innovations of The General Theory.

Jesper Jespersen untangles the concept and shows how Keynes, by taking uncertainty seriously, contributed to forming an analytical alternative to the prevailing mainstream general equilibrium framework:

Effective demand is one of the distinctive analytical concepts that Keynes developed in The General Theory. Demand and demand management have thereby come to represent one of the distinct trademarks of Keynesian macroeconomic theory and policy. It is not without reason that the central position of this concept has left the impression that Keynes’s macroeconomic model predominantly consists of theories for determining demand, while the supply side is neglected. From here it is a short step within a superficial interpretation to conclude that Keynes (and post-Keynesians) had ended up in a theoretical dead end, where macroeconomic development is exclusively determined by demand factors …

It is the behaviour of profit-seeking firms acting under the ontological condition of uncertainty that is at the centre of post-Keynesian concept of effective demand. It is entrepreneurs’ expectations with regard to demand and supply factor that determine their plans for output as a whole and by that the effective demand for labour.

Therefore, it was somewhat unfortunate that Keynes called his new analytical concept ‘effective demand’, which may have contributed to misleading generations of open minded macroeconomists to concluding that it was exclusively realized demand for consumer and investment goods that drives the macroeconomic development. Hereby a gateway for the IS/LM-model interpretation of effective demand was opened, where demand creates its own supply.

tmp10C1_thumb1On the contrary, it is the interaction between the sum of the individual firms’ sales expectations (aggregate demand) and their estimated production costs (aggregate supply) that together with a number of institutional conditions (bank credit, labour market organization, global competition and technology) determine the business sector decisions on output as a whole and employment …

The supply side in the goods market is an aggregate presentation of firms’ cost functions considered as a whole. It shows a relation between what Keynes called ‘supply price’, i.e. the sales proceeds that, given the production function and cost structures, is needed to ‘just make it worth the while of the entrepreneurs to give that employment’ (Keynes, 1936: 24). This means that behind the supply curve there is a combination of variable costs plus an expected profit at different levels of employment. At each level firms try to maximise their profit, if they succeed there is no (further) incentive for firms to change production or employment.

These assumptions entail that the aggregate supply function (what Keynes called the Z-curve) is upward sloping and represents the proceeds that has to be expected by the industry as a whole to make a certain employment ‘worth undertaken’ … In fact, this aggregate supply function looks like it was taken directly from a standard, neoclassical textbook, where decreasing marginal productivity of labour within the representative firm is assumed; the main difference is that Keynes is dealing with the aggregate sum of heterogeneous firms i.e. the industry as a whole.

The other equally important part of effective demand is aggregate demand function, which is the value of the sales that firms as a whole expect at different levels of macro-activity measured by employment (as a whole) …

Firms make a kind of survey-based expectation with regard to the most likely development in sales and proceeds in the nearer future. This expectation of aggregate demand (as a whole) is a useful point of departure for the individual firms when they have to form their specific expectation of future proceeds. This sales expectation will therefore centre around the future macroeconomic demand (and on the intensity of international competition).

Accordingly, Keynes’s macro-theory has a microeconomic foundation of firms trying to maximise profit, but differs from neoclassical theory by introducing uncertainty related to the future, which makes an explicit introduction of aggregate demand relevant i.e. the expected sales proceeds by business as a whole.

  1. October 26, 2020 at 2:45 pm

    “It is entrepreneurs’ expectations with regard to demand and supply factor that
    determine their plans for output as a whole and by that the effective demand for labour.
    Therefore, it was somewhat unfortunate that Keynes called his new analytical
    concept ‘effective demand’, which may have contributed to misleading generations of
    open minded macroeconomists to concluding that it was exclusively realized demand for
    consumer and investment goods that drives the macroeconomic development”.

    Helpful. Even more so, in his paper Jesperson adds:

    “Hereby a gateway for the IS/LM-model interpretation of effective demand was opened, where
    demand creates its own supply”.

    This, and Keynes’ explicit introduction of “aggregate demand”, is what I have been saying here for years about his adding “integral” feedback when the economy is treated not as equilibrium seeking in a Wheatstone Bridge but as (at the macro level) an information-based cybernetic or PID control system in which the primary control is through prices.

    Dipping into Kate Raworth’s “Doughnut Economics” to see how she treated this, I found in Chapter 3 a “circular flow diagram” in which three supplementary flows control the main flow.
    She says this was “for 70 years the defining depiction of the macroeconomy, so looked in my Samuelson (1964 international student’s edition) and found a somewhat similar multi-path diagram at Chapter 3, Figure 1). Of this Samuelson says much what I’ve been saying: “Twenty minutes poring over this diagram may be worth hours of disconnected musing about economic pricing”. Not being being able to display my own Wheatstone Bridge-type version of it perhaps explained why I get ignored, as sadly – and unjustly – it seems Professor Jesperson has been, so far.

    Perhaps more bloggers here will have Kate Raworth’s book, in which the “leakages” in that Figure 3/1 are clearly feedbacks. Fishing around for what she is now doing I found this:

    https://www.kateraworth.com/2012/07/23/why-its-time-to-vandalize-the-economic-textbooks/

    • October 26, 2020 at 2:49 pm

      Agologies for hidden newlines in copied text upsetting the layout intended above.

  2. Yoshinori Shiozawa
    November 9, 2020 at 10:07 am

    Effective demand is one of the distinctive analytical concepts that Keynes developed in The General Theory. (Jesper Jespersen)

    I agree with Jespersen that effective demand is one of the distinctive concepts that Keynes developed in The General Theory. However, I cannot understand why Lars Syll wanted to cite this part of Jespersen’s paper. Jespersen is cited by J. E. King (2015) Post Keynesian Economics as the author of “one of two excellent books on Post Keynesian methodology ” (p.35). It is true that Jespersen contends in his paper that this account of Effective Demand takes “the analytical procedure” that stand methodologically on Critical Realism (Capitals by Jespersen himself) but it is quite doubtful if his account of effective demand is really made standing on critical realism. In what sense is Jespersen’s account of effective demand based on critical realism?

    Tony Lawson, probably the main promoter of critical realism in economics, summarizes his idea as follows:

    Once, however, we change tack and give primary attention … to studying more directly the actual nature of social reality, a quite different and clearly more explanatory powerful or superior conception emerges. According to this social ontology, causality always matters and a more complex, processual account tends to dominates.

    The conception of social ontology I have in mind is processual in that social reality, which itself is an emergent phenomenon of human interaction, is recognized as being (…) highly transient, being reproduced and /or transformed through practice; essentially a process of cumulative causation (see Lawson, 2002a).
    (Lawson, 2016, What is neoclassical economics, p.36)

    Let us examine how Jespersen explained effective demand concept. He starts in contending that “Effective demand determines how much output and employment the business sector as a whole plan to undertake in the next ‘production period.'” ‘ (This sentence does not appear in the above citation. It appears between first and second paragraph that Lars Syll cited.)

    Then, at the fifth paragraph, Jespersen contends (or explains in his intention) that

    The supply side in the goods market is an aggregate presentation of firms’ cost functions considered as a whole. It shows a relation between what Keynes called ‘supply price’, i.e. the sales proceeds that, given the production function and cost structures, is needed to ‘just make it worth the while of the entrepreneurs to give that employment’ (Keynes, 1936: 24). This means that behind the supply curve there is a combination of variable costs plus an expected profit at different levels of employment. At each level firms try to maximise their profit, if they succeed there is no (further) incentive for firms to change production or employment.

    The last sentence means that, at each level of employment, there is a sale proceeds that each firm considers sufficient to produce and employ the given amount of employment. The aggregate production function Z = φ(N is defined on this reasoning. I know this is a loyal reproduction what Keynes gave in Chapter 3 in The General Theory. But is this justified in view of critical realist methodology? Is there any causal process in which the (aggregate) amount of labor employment determines (aggregate) sales proceeds? Readers and Jespersen are confused by Marshall’s concept of supply price, which Keynes reproduced in his book. There must be a causal process (or causality chains) in which (expected) sales proceeds determine the volume of production and hence employment, but it is the logic that defines employment function that Keynes argues in Chapter 21.

    Keynes is logically right, because he uses equilibrium framework and reasons in that framework. But, as is well known, no causality can be argued in an equilibrium framework. It supposes various constraints that would be satisfied when an equilibrium state is attained. No causality is implied. But, in view of social ontology to which Lawson and probably Jespersen subscribe, causality is important and in this argument direction of determination matters. A causal relation from (aggregate) expected sales proceeds to employment surely exists (if we do not ask accuracy of these relations). So, we can admit employment function but we cannot admit (aggregate) supply function. There is no such thing like aggregate supply function.

    A forma model of a monetary production economy is as well disastrous. Jespersen defines supply function is simply defined as

    Z_t = β(w, T, L_t ).

    The error of this definition is more manifest. Are there any definite relations between the total amount of employment when the real wage rate w and the technology T is given? Jespersen gives no explanation on this point. He simply assumes the existence of the function. It is easy to imagine two situations that have the same total employment where as total sales proceeds are different.

    Similar neglect of causality or processual account is observed in other places in Jespersen’s paper. This fact reveals that Jespersen’s methodology is not based on social ontology that Lawson has in his mind. Aggregate supply function is but an ill-defined concept.

    We must seriously question whether Jespersen’s account of effective demand and the significance of uncertainty on it. Lars Syll must share some part of responsibility to have cited Jespersen’s paper approvingly.

    • November 10, 2020 at 12:02 pm

      We must seriously question whether Yoshinori understands the difference between a mathematical constant and a variable if he thinks Tony Lawson’s ever-evolving social ontology could be anything but ill-defined, as against differentiated from the meanings of other terms. Causality, like evolution, is a way of accounting for change, and so is the Labour in Jespersen’s supply function. It doesn’t account for how much the Labour produces, it indicates a space whose content needs to be looked at and evaluated in particular cases.

      Since Yoshinori may have ignored the passage of Jespersen which I found helpful because of my quotation of it having become ill-formatted, let me repeat it:

      “Therefore, it was somewhat unfortunate that Keynes called his new analytical concept ‘effective demand’, which may have contributed to misleading generations of open minded macroeconomists to concluding that it was exclusively realized demand for consumer and investment goods that drives the macroeconomic development”.

      In other words, Keynes’ concept (an unset variable) could easily be mistaken for a value by those more familiar with arithmetic than algebraic structure.

    • Yoshinori Shiozawa
      November 11, 2020 at 12:54 am

      Addendum: The last paragraph in my post above on November 9, 2020 at 10:07 am contained an incomplete sentence. The first sentence must be read as follows:

      We must seriously question whether Jespersen’s account of effective demand and the significance of uncertainty on it are justified in view of critical realism.

  3. Yoshinori Shiozawa
    November 11, 2020 at 1:38 am

    Addendum 2: The paragraph that starts with “A forma model of a monetary production economy” also contains an incomplete sentence. The second sentence must be read as follows:

    Are there any definite relations in which the total amount of employment (when the real wage rate w and the technology T is given) leads to the total sales proceeds?

    There are some other grammatical errors but I do not correct them. Readers can understand what I wanted to say.

  4. jesper jespersen
    November 12, 2020 at 6:33 pm

    First of all thank you to Lars Pålsson Syll, Yoshinori Shiozawa and Dave Taylor for taking up this issue on ‘How to understand and make use of Keynes’s crucial/significant novelty ‘the principle of effective demand’. Probably the most creative contribution to macroeconomics in the 20th century. Here, we are fortunately in agreement.
    Having said that, therefore, it is so important that this concept is debated among serious scholars, because Keynes is not that easy to interpret, and as it was a genuine novelty, he had to develop his own language on this issue – (although partly relying on the tradition at his time – just to be (partly) understood, or rather misunderstood, by his contemporary colleagues).
    Then Yoshinori Shiozawa raises the question, whether my interpretation of Keynes’s Principle of effective demand is grounded in the philosophy of science called ‘Critical Realism’ and he quotes Tony Lawson rather lengthly to emphasise the importance of dealing with a’ ‘social ontology, [where] causality always matters and a more complex, processual account tends to dominates’.
    Once again, I could not be any more in agreement: the social ontology we are dealing with is complex and partly hidden, hence suggested causalities are unavoidable for a better understanding of the working of the macroeconomic system. One of the epistemological problems is that these (suggested) causal relations can only partly, if at all be (empirically) tested – and to the truth we will never get (Keynes and Popper).
    Let me take issue with another of Yoshinori Shiozawa (YS) sentences: Keynes is logically right, because he uses equilibrium framework and reasons in that framework. Of course, Keynes was logically right educated in Mathematics and Philosophy. But, YS are mislead by the word ‘equilibrium’ used by Keynes in a different analytical meaning than neoclassical macroeconomics (read my chapter 6 in Macroeconomic Methodology). A possible interpretation is a ‘stand still’, which has to be explained by causal relations in an ever evolving macroeconomic environment. This is what call an analytical semi-closure, which gives birthing space to understand part of the dynamic causal related macroeconomic system. (Keynes use of this concept ‘equilibrium’ was of course misunderstood by his neoclassical opponents – among others Arthur Pigou).
    Whether there is a causal dynamic connection from changes in aggregate employment to expected demand is to me rather likely taking Keynes’s discussion of the multiplier effect into consideration, which he (with good inspiration from Ricard Kahn and Jens Warming) took on board as another important dynamic novelty.
    At the end of Yoshinori Shiozawa’s comment, he concludes:
    ‘We must seriously question whether Jespersen’s account of effective demand and the significance of uncertainty on it are justified in view of critical realism’.
    I have to say, that I have got so much inspiration from following the Critical Realist weekly workshop in Cambridge during several terms, so I am not in doubt that from a point of view of philosophy of science I am in deep debt to this group lead by Tony Lawson (and to Victoria Chick and Roy Rotheim). This came out in my dissertation on Macroeconomic Methodology, Edward Elgar 2009.
    Having said that, it also came clear to me participating in this workshop on Critical Realism is that CR is a building with many rooms: what room you occupy depends on the specific case and context. And therefore, I am pleased to conclude this comment by emphasising have much all four of us are in agreement of giving the ‘Principle of effective demand’ an even more important role in Keynes’s macroeconomics than usually is the case even by honest Keynes scholars.

    • Yoshinori Shiozawa
      November 18, 2020 at 5:54 am

      An explanation would be necessary for new comers. The above post of jesper jespersen dated on November 12, 2020 at 6:33 pm did not appeared on the screen by what reason we do not know. So, Jespersen posted his “Reply” below dated on November 14, 2020 at 6:05 pm and Yoshnori Shiozawa and davetaylor1 posted their “Replies” on the assumption that jespersen’s post November 12, 2020 was lost. Then, the lost post appeared on the screen and jespersen wrote his post on November 15, 2020 at 5:51 pm.

  5. jesper jespersen
    November 14, 2020 at 6:05 pm

    Dear Yoshinori Shiozawa,
    I wrote you a long reply on Tuesday, which might have disappeared. In that case I will be back with a number of substantial comments, because your writing ask important questions to be answered
    Sincerely

    Jesper Jespersen
    Professor of economics, Roskilde University, Denmark

    • Yoshinori Shiozawa
      November 15, 2020 at 3:15 am

      I am sorry that you have lost your post, especially when it was a long one. I have experienced the same misfortune at least two times. I learned that it is necessary to copy the text after having written at first on another file if it is long. Don’t hurry too much. I can wait.

  6. November 14, 2020 at 9:44 pm

    Since Professor Jespersen and I have corresponded privately, I may be able to help in reconstructing his “long reply”.

    He sent me a paper, the title of which quotes Keynes: “Look after employment; the budget will look after itself”. This most helpfully explains:

    “‘To escape the old ideas’ meant giving up the pre-condition that the macroeconomic development is supply-side determined. … Keynes’s new idea is simply to incorporate the demand side into the macroeconomic system. Employment depends on production (Keynes called it ‘output’) and production depends on

    1. Expected demand for goods and services and
    2. Profitability.

    “It is the combination of 1. and 2. which Keynes called ‘the Principle of Effective Demand’, see (Jespersen 2012). The entire 1936-book could be read as one long theoretical argument explaining the factors which make an impact on effective demand, and how they are interrelated within a macroeconomic system, what Keynes called ‘The economy as a whole.’”

    Thanking Jesper for this, I summed up the argument thus. “Yoshinori has stuck seeing Keynes saying ‘Savings equal investment’ – rather than ‘Savings equal REAL investment’, which discounts savings tied up in speculation.”

    Discussion of the speculation needs to take into account that Keynes’ theory of subjective expectations (probabilities), though relevant, was notoriously different from the supposedly objective theory in which subjective classification is taken for granted by econometricians.

  7. Yoshinori Shiozawa
    November 15, 2020 at 3:42 am

    Sorry, Dave, your comment is off the mark. I am asking Jespersen how he can justify the causal relation such as “Z = φ(N)” or more precisely how one can claim there exists any causal relation or process from the number of employment to the sales proceeds. I am also asking him how he can claim that a state defined by posing Z = D (equilibrium) is causally determined in any sense.

    When talking causal relations, you cannot simply take the inverse function instead of employment function..Even if an equilibrium state exists, that equilibrium is not a state that is causally induced or results by/from some causal processes.

  8. jesper jespersen
    November 15, 2020 at 5:51 pm

    I am so happy, Edward found my long comment from Tuesday, which should clear the points originally raised by Yoshinori Shiozawa (YS). So far so good, hopefully.

    So, I will be brief.
    Unfortunately Yoshinori, you are mistaken in your rigid interpretation of ‘equilibrium’. Keynes (and I) are arguing within a dynamic simultaneous and open macroeconomic system – path-dependent (if you like), which (only) occasionally might end-up in a ‘stand still’ with substantial unemployment – this is what Keynes called an ‘equilibrium’ in the lack of a better word; but so misleading to his neoclassical colleague (and still is – even today).
    This ‘stand still’ is characterized by D ==>Z (given a number of conditions, which I (echoing General Theory, chapter 18, p.243) discuss in my paper on ‘Effective Demand’ and my book on ‘Macroeconomic Methodology’)!
    But if anything D ==> Z is a causal relation with in Keynes’s model.

    • November 16, 2020 at 7:55 am

      “Keynes (and I) are arguing within a dynamic simultaneous and open macroeconomic system …”.

      As am I. Steering (cybernetic control) AIMS for a dynamic equilibrium (staying more or less on course) in order to finally achieve stationary equilibrium (an intended port). It doesn’t stay on course unless the steersman CAUSES it to; it will go way off course if (receiving differential feedback) he changes course to avoid on-coming traffic but doesn’t think to reset it. (Chaos)!

      In Algol 68 this is expressed simply by replacing Yoshinori’s Fortran “=” by ” Z” with the order inverted so “Z <= D", where D can be the result of a process. For those not computer-old enough to remember, 'Fortran' is an abbreviation for "formula translator", the original mathematical programming language

      So thanks for this, Jesper. Your clarification has been greatly appreciated..

      • November 16, 2020 at 8:03 am

        This was definitely not a typo. Where ” Z” has been received I sent Jesper’s expression ‘C++-like “D ==> Z”‘..

      • November 16, 2020 at 8:13 am

        To explain “<=" I also sent '(read "becomes equal")'. I didn't get round to saying Jesper's version is often read "maps into". Really, though, I'm not surprised that on this site Jesper's earlier comment got lost.

      • November 16, 2020 at 12:25 pm

        It seems Craig had the same problem when writing about Limits of Mainstream Economics at 3.15 a.m. on 15 November 2020. Yoshinori too has has had problems with what I thought were typos but Jesper suggests may be a dodgy spam filter. Something the editor needs to look into, anyway. No wonder we can’t agree with each other when when we’re not getting the whole story.

        Wondering how far Yoshinori was right to accuse me of interpreting words by their sounds (my counter being that in scientific contexts words are often coined on that basis), I looked up “retroduction” (which I’ve read as the opposite to “deduction”) in Hartwig’s “Dictionary of Critical Realism”. There I was referred to a Bhaskarian discussion of what Craig was talking about: Dialectical argument. The following passage is a pretty fair summary of what I had seen “for real” in my own work a good twenty years before finding it “on paper” in Bhaskar’s “Dialectic: the Pulse of Freedom”. Says the dictionary:

        “A dialectical argument or dialectical explanation is a dialectical argument that substantively explains ‘contradictory forces in terms of a structured common ground’. …

        “”In the preface to ‘Dialectic’, Bhaskar describes his work as involving ‘a preservative and enrichment of hitherto existing critical realisms that is ‘a non-preservative sublation of Hegelian dialectic’. To Hegel’s [Craig’s?] dialectic of identity, negativity and totality, Bhaskar will offer the four terms of non-identity, negativity, totality and transformative agency, and he adds that where he and Hegel share common terms DCR will provide radically different interpretations …

        “The dialecticisation of CR is itself a dialectical PROCESS that has four moments (see MELD). Bhaskar schematises this theoretical development as from a first moment (1M) to a second edge (2E) then to a third level (3L) and on to a fourth dimension (4D). [Representation by points, lines, planes and a sphere expanding in time?] IM involves existing CR concepts like structure, differentiation, change, alterity (for example epistemic/ontic non-identity within ontology), transfactual efficacy , EMERGENCE and systemic openness. …”.

        I can go along with this, though I can understand Yoshinori’s distaste for the language of it. I find the four levels more obviously explained in terms of the use of the four types of data provided for in Algol68. [Constants, variables, their real modes of interpretation and the procedures which realise the variation]. The Algol68 compiler is itself an Algol68 program.

    • Yoshinori Shiozawa
      November 18, 2020 at 8:08 am

      As Jesper jespersen’s lost post on November 12, 2020 at 6:33 pm was retrieved, we are all happy. Jespersen had a good luck. Now, we can begin our interrupted discussion.

      I have ordered your book (2009) Macroeconomic Methodology: A Post-Keynesian Perspective. It will be my pleasure to read Chapter 8 Effective demand: a macroeconomic causal relationship. Unfortunately, the copy arrives at my place in late December (probably after Christmas). So, let me ask some questions without waiting the end of the year.

      You (jespersen) wrote in the post on November 12, 2020 at 6:33 pm that

      YS [Yoshinori Shiozawa] are [sic] mislead by the word ‘equilibrium’ used by Keynes in a different analytical meaning than neoclassical macroeconomics.

      I want to ask you only thee questions this time to avoid misunderstanding between us.

      (1) Would kindly explain concisely what you mean by “a different analytical meaning than neoclassical macroeconomics”?

      (2) The point of intersection of D-curve and Z-curve, which you called point of effective demand, is it an equilibrium point or any other point for you? How do you characterize this point in your own terms?

      (3) Keynes wrote (in Chapter 3 in The General Theory) that, at the equilibrium level of employment i.e. the point of intersection of two curves, “there is no inducement to employers as a whole either to expand or to contract employment.” In neoclassical economics, too, an equilibrium state is defined as the point point at which agents (firms or consumers, depending to the situation we observe) have no incentive to change their behavior. How does Keynes’s equilibrium level of employment differ from the meaning of “equilibrium” in neoclassical economics?

      Please note that I am not asking in question (3) the difference between employment equilibrium (Keynes) and price equilibrium (in the standard model of neoclassical economics). I want to know how you differentiate the theory-structural difference between the two.

  1. No trackbacks yet.

Leave a Reply

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

WordPress.com Logo

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

Google photo

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

Twitter picture

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

Facebook photo

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

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.