Skidelsky on the uselessness of ‘New Keynesian’ economics
from Lars Syll
Whereas the Great Depression of the 1930s produced Keynesian economics, and the stagflation of the 1970s produced Milton Friedman’s monetarism, the Great Recession has produced no similar intellectual shift.
This is deeply depressing to young students of economics, who hoped for a suitably challenging response from the profession. Why has there been none?
Krugman’s answer is typically ingenious: the old macroeconomics was, as the saying goes, “good enough for government work” … Krugman is a New Keynesian, and his essay was intended to show that the Great Recession vindicated standard New Keynesian models. But there are serious problems with Krugman’s narrative …
The New Keynesian models did not offer a sufficient basis for maintaining Keynesian policies once the economic emergency had been overcome, they were quickly abandoned …
The problem for New Keynesian macroeconomists is that they fail to acknowledge radical uncertainty in their models, leaving them without any theory of what to do in good times in order to avoid the bad times. Their focus on nominal wage and price rigidities implies that if these factors were absent, equilibrium would readily be achieved …
Without acknowledgement of uncertainty, saltwater economics is bound to collapse into its freshwater counterpart. New Keynesian “tweaking” will create limited political space for intervention, but not nearly enough to do a proper job.
Skidelsky’s article shows why we all ought to be sceptical of the pretences and aspirations of ‘New Keynesian’ macroeconomics. So far it has been impossible to see that it has yielded very much in terms of realist and relevant economic knowledge. And — as if that wasn’t enough — there’s nothing new or Keynesian about it!
‘New Keynesianism’ doesn’t have its roots in Keynes. It has its intellectual roots in Paul Samuelson’s ill-founded ‘neoclassical synthesis’ project, whereby he thought he could save the ‘classical’ view of the market economy as a (long-run) self-regulating market-clearing equilibrium mechanism, by adding some (short-run) frictions and rigidities in the form of sticky wages and prices.
But — putting a sticky-price lipstick on the ‘classical’ pig sure won’t do. The ‘New Keynesian’ pig is still neither Keynesian nor new.
The rather one-sided emphasis of usefulness and its concomitant instrumentalist justification cannot hide that ‘New Keynesians’ cannot give supportive evidence for their considering it fruitful to analyze macroeconomic structures and events as the aggregated result of optimizing representative actors. After having analyzed some of its ontological and epistemological foundations, yours truly cannot but conclude that ‘New Keynesian’ macroeconomics, on the whole, has not delivered anything else than ‘as if’ unreal and irrelevant models.
The purported strength of New Classical and ‘New Keynesian’ macroeconomics is that they have firm anchorage in preference-based microeconomics, and especially the decisions taken by inter-temporal utility maximizing ‘forward-looking’ individuals.
To some of us, however, this has come at too high a price. The almost quasi-religious insistence that macroeconomics has to have microfoundations – without ever presenting neither ontological nor epistemological justifications for this claim — has put a blind eye to the weakness of the whole enterprise of trying to depict a complex economy based on an all-embracing representative actor equipped with superhuman knowledge, forecasting abilities and forward-looking rational expectations. It is as if these economists want to resurrect the omniscient Walrasian auctioneer in the form of all-knowing representative actors equipped with rational expectations and assumed to somehow know the true structure of our model of the world.
And then, of course, there is that weird view on unemployment that makes you wonder on which planet those ‘New Keynesians’ live …
Leave a comment
—– look inside —– $5.94 / $20.00
—– look inside —– $4.90 / $8.00
—– look inside —– $15.99
—– look inside —– $5.99 / 12.99
—– look inside —– $5.93 / $12.99
—– look inside —– $4.97 / $9.90
—— Ugarteche, Puyana and Madi ——
Gerson Lima / Maria Alejandra Madi
Edward Fullbrook and Jamie Morgan
————— Michael Hudson ————–
Maria Alejandra Madi / Jack Reardon
————- Edward Fullbrook ————-
—————— Steve Keen —————–
————— Richard Smith —————
————– Gustavo Marques————
– Victor Beker and Beniamino Moro –
————– Lars Pålsson Syll ————-
—————– Stuart Birks —————-
Edward Fullbrook and Jamie Morgan
———— Armando Ochangco ———-
Shimshon Bichler / Jonathan Nitzan
————— Mauro Gallegati ————–
————— Herman Daly —————-
————— Asad Zaman —————
—————– C. T. Kurien —————
————— Robert Locke —————-
Real-World Economics Review Blog 
Why pick out New Keynesian economics to castigate? There is NO valid theoretical quantitative orthodox or heterodox analysis. All mathematical forms fail the requirements of the quantity calculus. That no guesses of mathematical form have been found for any possible abstract theoretical equation, which is not invalidated by the empirical evidence should have demonstrated the sheer futility of this approach. Only by starting from first principles will valid equations be developed.
Because the promotion of facile New Keynesianism has taken the eyes of successive generations of trainee economists off the real Keynesianism, preventing it being understood and its obscurities ironed out. As I’ve been saying since 1968, in correcting employment as well as price levels, Keynes was anticipating cybernetic control theories not articulated until 1948. I had worked with paradigmatic examples of such systems in 1960-1 in which the terms of the theory were visibly represented by distinct components of the system. By 1968 the significance of time delays turning negative into positive feedback had been understood. I only learned c.2010 that it its feedback circuits had been reduced to a microchip with the mathematically explicit name PID servo. Where in theoretical terms Keynes had advanced to PI, the New Keynesians ignored the I (Keynes’ unemployment) and taught only P (pricing).
While I am sympathetic to your argument, Frank, in effect I am denying your assertion that no mathematical form has been found which has not been invalidated by the evidence. That depends on what you mean by ‘mathematics’. Continuous one-dimensional real numbers and dimensionless quantities are ontologically subsets of two-dimensional complex numbers, the generic form, even though epistemically the complex was constructed from the simpler forms. PID can be portrayed as a complex number (a cross roads); a real number can represent only a single process (a line) with no side effects (hence equilibrium).
Apologies if this is over the heads of specialists from non-mathematical disciplines, but it (along with Fullbrook’s inverting the measure of Market-Value) is fundamental to seeing and understanding the way past the outdatedness of Samuelson and Krugman’s neo-liberal New Keynesianism. The Fullbrook argument, and Levi-Strauss’s argument for the appropriate representative agent being the family [a complex system] rather than quantifiable individuals, were kicked around less mathematically in my recent comment on economic philosophy:
All current economic theories are based on micro-foundations as the link to a Steve Keen video below shows. He suggests deriving macro only from macro. That’s fine, but then why are we not deciphering the current monetary and financial paradigm and then trying to find a new paradigm to replace it?
Paradigms being both the conceptual essence of a body of knowledge/area of human endeavor and its temporal effects IS THE QUINTESSENTIAL MACRO PERSPECTIVE, NO?
In fact the title of the last power point slide in the above video is entitled Building a New Paradigm in Economics.
The problem of course is Keen and no one else has discovered the new economic insight of the integrative problem resolving power of a high percentage discount/rebate monetary policy at the point of retail sale, which is also the very expression of the new paradigm of Direct and Reciprocal Monetary Gifting, and so progress flounders at best and at worst insures that we run lemming-like over the cliff of economic collapse and ecological suicide.
To paraphrase Richard III’s lament: “An insight, an insight, the planet for an insight.”
Dave my reply to your comment is below at Nov 23, 9.09hr.
Keen (are you listening Steve) should model monetary gifting directly to the individual in the form of a universal dividend and directly and reciprocally at the point of retail sale in Minsky and in his Godley tables and as new reserve accounts at the central bank labeled Dividend Distributions and for retail enterprise labeled Discounts/Rebates and see how the policies of a universal dividend and a 50% discount/rebate policies at retail sale resolve our deepest economic problems.
6-7 years ago when I was posting to his blog about the importance of examining the economy and money system from a cost accounting perspective he suddenly realized the importance of double entry bookkeeping and the fact that economists could get their PhD and never have to take so much as a rudimentary class in accounting.
Understanding accounting is the ultimate grounding discipline for the abstract mathematical and theoretical fugues that economists and pundits fall into by helping them to focus on the actual temporal process of the economy where they might cognite on the significance of the ending, summing and terminal expression point of retail sale (it’s where production becomes consumption after all) and for economic factors like cost, price and inflation, and how a monetary policy at that point is so consequently powerful.
Problems are complex. Solutions, particularly integrative ones like paradigm changes are always simple, elemental and permanently progressive.
I concur with Lars criticism of any macroeconomics that uses representative agent theorising. Aggregation theory shows that a representative agent does not exist unless all actual agents are identical; there are no weaker conditions for a reasonable approximation either. Telling choice theoretic fairy stories about a non-existent entity does not constitute “microfoundations” and I am at a loss to explain the popularity of this approach.
Apart from relying on empirical regularities corresponding to common-sense conjectures (eg people with more money buy more goods), the best way to proceed seems to be by so-called agent-based models. Behaviour of disparate agents can be based on empirical studies and extensive simulation used to see under what conditions aggregate data patterns are reproduced. There are obvious risks and criticisms of that but we can’t do better than the best we can.
You are absolutely correct in supporting agent based models. The major problem in using such an approach is that each result is only a single numerical solution. It is necessary to map the whole solution space. I know this to be true because this is how I started my analysis which led me to realise what was significant. Certain regularities implied that an analytical representation existed. This formulation is presented in my paper, Transient Development, RWER-81. The abstract theory of production it presents is fully in accord of every known relationship found in economic analysis and it is not invalidated by the empirical analysis.
Frank, let’s try an example to see if we can agree.
It seems to be true that people whose income goes up buy more stuff. We don’t know what stuff and we don’t know how long they spend shopping – those data are not officially collected. We know people’s incomes in total, roughly, from income tax data and we know their consumption expenditure in aggregate from value-added or sales tax receipts. The national Stats agency checks the total by sample surveys of households – how much they earn, how much they spend. Now, here’s the economist confronted by two official data series, measured in pounds, dollars or Euros depending on the country: aggregate income of households, aggregate expenditure of households on consumption goods. He asks: I wonder if the relation between these two has been stable and can be used for conditional forecasting? He writes down an equation and estimates its parameters on the data. He checks for functional form by non-nested statistical tests and picks the best (it need not be linear but he keeps it as parsimonious as he can), He checks for historical stability of the relation – various techniques. He studies the error terms and he adds various conditioning variables on the RHS all measured in monetary terms or else dimensionless (like interest rates). He can’t add irrelevancies; there has to be a story as to why the variable might influence consumption decisions but he lets the data determine whether it has or not. If he is lucky he comes up with a “consumption function”. it does not solve most of the problems that worry people on this blog but if it has been stable in the past it will probably be stable in the future until a wholly new circumstance arrives to make an omitted variable relevant. The economy displays considerable inertia.
His equation relates monetary quantities to monetary quantities. It says nothing about physical quantities and it is not a “law”. In what sense does it fail the requirements of a quantity calculus? What would you have the economist do?
It fails the quantity calculus in the form of the equations used. If you follow examples of what you have described you will find that the equations for consumption follow the equations used in production functions such as the Cobb-Douglas equation. Money is a quantity! What is wrong with all of them is that they raise quantities to fractional powers. This is an error. Every theoretically valid equation can only contain quantities raised to small integer powers. If you look at equations (14), (15), (23) and (24) in my paper you can see that is true. The exponential terms are of numbers of dimension one which are not quantities. To summarise, these equations can NOT have theoretical validity, They are merely fitted equations. Any arbitrary equation can be used.
I would have economists practise the scientific method which demonstrates that all conventional quantitative analysis can NOT be valid. Then they need to start again and apply first principles analysis so they can determine valid quantitative relationships.
As a matter of fact most production functions are log linear or linear and don’t use power terms at all, fractional or integer. So does that mean they are oK in your world.?
Don’t get hung up on the Cobb Douglas function. It started ;life because of a belief in a “law” of diminishing returns or variable proportions, which the function reflects. That might hold for a particular process but even so there is no reason why an aggregate function would have the same form. As an aggregate function it has no particular justification. Your model by assuming strict complementarity between present and past or maintenance labour time and ignoring the possibility of substitutability is a special case. A more general formulation would run into the same aggregation issues as Cobb Douglas. The notion of an aggregate production function is highly contested in economic theory.
As I have said previously, all production functions are theoretically invalid. Logarithms can only be applied to pure numbers. Therefore logarithms are excluded by the quantity calculus.
You are misunderstanding my analysis. It takes into account the possible substitution between labour and capital. If you look at my figures (1a), (1b), (3a) and (3b), you can see that they range from h_d being zero to one. That is no capital increasing to only creating capital. The y-coordinate shows the change in output for all possible substitutions.
My relationships are the general case — definitely not a special case. All the other forms you mention are special cases tied to individual data sets.
The reasons there is a problem with the idea of an aggregate production function is the misunderstanding that aggregation is not possible although all the empirical evidence suggests it is. The reason is that the aggregation of labour-time is valid. Then if the monetary values used as a proxy for output are an affine transformation of labour-time, then aggregation appears meaningful. Hence macro-economists are pulled in two directions. It appears to work empirically but they can not justify it theoretically. But once the labour-time relationship is understood, there is no conflict.
It would have been better If I has finished the first line as invalid as a theoretical relationship.
The poster above represents a pig is that is “still neither Keynesian nor new”. I believe the poor pig would have been to intelligent to wear such bright red lipstick. I am sure we have all read Animal Farm. Time to develop a new agenda based on the calamitous issues that face us globally?
Palliative theoretics is palliative. Pattern changes are transformative.
Isn’t it time then that we considered the signatures of new paradigms and then began an examination of the singular concept that paradigms are and whose aligned policies create its temporal universe effects?
What is this obsession with theory testing all the time? There is no theory in the hypothesis sense, just the practical application of all the existing laws that are present, such as the Constitution giving the right to create currency to the national sovereign government. Plus the various Currency laws[in the USA] from 1792 which talked about the dollar as the currency medium, Only some screwed up theory started the idea a government had to tax and borrow or save it’s currency to allow it to spend. Even though it’s been very pervasive, it was always wrong. Academia gone berserk.
John, the “tax or borrow to spend” idea is not academia gone beserk but a very hard-headed arrangement by which bankers in London established control over the activities of the new “representative” English government and constitutional monarchy following the “glorious revolution” of 1688, which replaced our hereditory king. The writers of your US constitution fled England to avoid this, hence their claiming the right to create their own currency. The subsequent re-establishment of the English system in the Federal Reserve Bank of 1913 was by fraud, according to “The Money Masters” documentary and other sources. You will understand the US better if you study English, Scots and Irish history and the formation of the UK since just before the Reformation. Book 1 of More’s ‘Utopia’ is an excellent place to start, for it discusses the results of the mass production of wool – which Adam Smith generalised.
Hi Dave. The history of economics can explain the origins of some of the mess we see today. I don’t doubt that tax to spend was a bad idea when the concept of fiat money was I believe unknown.Money itself had intrinsic value and was used by the state,not created by the state. It doesn’t mean we have to do it today.
Dave,
Thank you for being sympathetic with my argument but the argument you are presenting makes a category error. Physical quantities describe tangible elements. Therefore they can not be represented by some mathematical form which is physically impossible. So the logarithm, exponential, fractional power and complex numerical forms are impossible when applied to abstract (that is theoretically valid) representations of quantities. This is all the quantity calculus says. Can you imagine the square root of a length? One can easily see the square root of an area and the cube root of a volume are lengths. What is the complex part of a line? It can only be zero.
I said and continue to say that no orthodox nor heterodox formulation representing production has been published. All continue to include forms which the quantity calculus forbids. The only possible representation of production is in my paper Transient Development.
Frank, you are being too defensive!
I am sympathetic with your position on the quantitative calculus because Gibbs’s extension of the real dimensions into an indefinite number of formal ones ignores the rounding errors inherent in the arabic number format: evident, as I said, in its incommensurability with pi, but even even more so in scientific number notation (as in 10 to the n times a limited number of decimal places). It is not that the scientific notation is (as you suggest) impossible, it is just that it is being applied to the tangible form of approximating number formats rather than the form of the real universe, expanding in time.
We are at cross purposes on economics. Your theory is about production, mine (as Ted has helped me to see) is about bridge building between economists and information science, as well as a topological mapping of the evolution and current form of the human economy. What I have learned from you is that in the present day world, with its environmental crisis and vast data processing resources, we should be accounting not just for money usage but for usage of specific real resources. That is what your diagrams seem to be indicating.
I did not intend to imply that scientific notation is impossible just that standard economists are barking up the wrong tree. What is required is first principles analysis — the normal approach in the physical sciences. I am not sure what you mean in discussing methods of representing numbers and their only providing an incomplete set. As far as mathematics is concerned there are an infinite set of numbers between any two points on the number line however close together they are.
Frank, you said: “Physical quantities describe tangible elements. Therefore they can not be represented by some mathematical form which is physically impossible. So the logarithm, exponential, fractional power and complex numerical forms are impossible when applied to abstract (that is theoretically valid) representations of quantities.”
Representations of forms ARE tangible elements, which is WHY scientific notation is possible! In your latest comment, I agree on economists with the probable exception of Leontiev. Your “first principles” analysis applies to mathematics as well as science. My discussion of incompleteness was trying to draw attention to circular “number lines” being incompatible with linear ones, for even if both theoretically involve “an infinite set of numbers”, in practice there not an infinite number of digits with which to represent them.
I’m coming at this from the use of complex numbers in electrical and communication engineering, where they are used in managing timing differences and using these to convey information. What I’ve been struggling with seems to be your “abstract representations of quantities”, for the point of representing something is translate an abstraction into something tangible. I think I have now grasped the idea of a “real quantity” being an abstraction from a quantity of something real (e.g. peanuts); which is fair enough. Your talk of a “number line”, however, suggests you haven’t understood my explanation of complex number.
Googling “complex number” in hope of finding a link to a clear explanation of it, I was appalled by the confusion and ignorance I found. No wonder you haven’t understood it, Frank! In the end I found an interesting history, but what I was looking for was under not “complex number” but “complex plane”. Put simply, there are algebraic and geometric ways of expressing complex numbers, but in the digital age the now usual algebraic ordered pair (a,b) format leaves out the operator i, which in the geometric form rotates a number line through a right angle, as in Pythagoras’s original square. The (a,b) form invites the mathematical economist’s indefinite extrapolation to n-tuples (a,b,c,d, …. n) and matrix algebra, whereas, I can now clearly agree, though in our real universe there are only four dimensions to abstract from, yet everything in it can be located dimensionally in space and time. From the point of view of the tangibility of the resultant format, I’ve previously called the algebraic form “symbolic” and the geometric form “iconic” (from economist Kenneth Boulding’s delightful book “The Image”).
https://en.wikipedia.org/wiki/Complex_number#History
https://en.wikipedia.org/wiki/Complex_plane#Argand_diagram
What I was trying to say about complex numbers being the generic form and simple quantities a particular case is summed up in this conclusion:
“Work on the problem of general polynomials ultimately led to the fundamental theorem of algebra, which shows that with complex numbers, a solution exists to every polynomial equation of degree one or higher”.
Dave, I really do understand complex numbers. I was trying to explain that quantities can only be added when they are the same type of quantity and they can be multiplied and divided without restriction. The ability to ascribe specific values is irrelevant. For example we use pi to represent a number which is impossible to represent in any other way.
Calculus deals the curves of any shape.
When I said tangible forms, I meant the physical unit must be seen to exist. For example the Cobb-Douglas equation with its power terms manipulates values which can not be quantities only a numerical values.