Limits of marginal productivity theory

It’s fitting that Mervyn King should have been  a director at Aston Villa – because that job, like his previous one, has given him a close-up view of the failure of mainstream economics.

I’m referring to the idea that wages equal marginal product. Aston Villa’s wage bill this season was higher than Leicester’s. This difference is not reflected in their marginal product. In fact, Villa owner Randy Lerner is sitting on a loss of over £200m.

This highlights an under-appreciated distinction: people are often not paid their actual marginal product, but their expected marginal product. Villa paid Joleon Lescott and a ragbag of Carlos Kickaballs £50,000 a week because they expected them to save the club from relegation. Such expectations proved over-optimistic.

This is not an idiosyncratic example. Across whole professions, wages can be inflated because expected marginal product exceeds actual marginal product. Fund managers, for example, are paid millions for mediocre performance because gullible investors over-estimate their ability to beat the market. I suspect the same is true of corporate CEOs: managerialist ideology causes remuneration committees to over-estimate the extent to which an heroic leader can improve the company, and so pay too much. If bosses were paid their actual marginal product, Fred Goodwin and Dick Fuld would have had salaries of minus billions of pounds. They didn’t.

I’d add that there are at least three other reasons why pay can deviate from marginal product, especially at the top end.

First, value is often determined not merely by individual’s production, but by team production, which makes it difficult to identify an individual’s marginal product. As Lars Syll says (pdf):

It is impossible to separate out what is the marginal contribution of any factor of production. The hypothetical ceteris paribus addition of only one factor in a production process is often heard of in textbooks, but never seen in reality.

For example, the same manager, for example, might or might not add value depending (pdf) upon whether his skills are a good match or not with the organizations.

This can cause pay to exceed marginal product because the fundamental attribution error leads hirers to over-estimate individuals’ contributions and under-estimate organizational factors – as happened, for example, when ITV hired Adrian Chiles and Christine Bleakley on big money in the mistaken hope that they would attract viewers.

Secondly. wages can exceed marginal product in efficiency wage situations, where workers must be bribed not to steal the firms’ assets. This explains why bankers are paid so much. I suspect the same applies to CEOs.

Thirdly, power matters. Dani Rodrik has shown that even controlling for productivity workers are better paid in democracies (pdf) because these are associated with stronger workers’ bargaining power. In the same spirit, economists at the IMF show that CEOs pay is lower when trades unions are stronger.

All of this is to vindicate a point made by Joe Stiglitz – that evidence for the validity of marginal product theory, especially at the top end, “remains thin”.

Now, you might object that some of the examples I’ve given are of disequilibrium: Fred Goodwin’s job didn’t last long, and Villa players’ prospects aren’t great. But all theories are true if you ignore the exceptions. Life is lived in disequilibrium, and a few years of egregious deviations from marginal product can generate big inequalities.

It could be that marginal product theory – just like simple-minded talk of incentives – is as much ideology as science.

Comments

[TickyW]

Heh! Great stuff, Chris.

[Luis Enrique]

I think it's a bit much to call this a failure of mainstream economics. To pick one paper that googling marginal product in teams produces, Alchian and Demsetz was published in the AER in 1972 and has 15143 citations in google scholar. The idea that marginal products are not defined or more complicated, or are probabilistic etc. are very mainstream. Kremer O-ring production functions, QJE 1993, can't get more mainstream than that.

Suppose you take a random sample of 100 mainstream economists and ask them what determines wages, how many of them will say something like "well, a simple starting point is concept of marginal product, idea you're not going to pay somebody more than they are worth to your business, and that if markets competitive they will get what they are worth .. but then labour markets are very far from competitive, in many situations marginal product not meaningful concept, there are various frictions, rent extraction at the top end, it's all really about bargaining power.... etc."

my guess is upwards of 80 would say something like that. I spend my life reading characterisations of mainstream economic on the internet that bear very little relation to the mainstream economists of my experience. There is a cottage industry of generating and maintaining religious beliefs about how dreadful mainstream economists are.

and I know I have said this a thousand times already, and I really don't know why you aren't updating your ideas on this, maybe capitals will help, but THE MAINSTREAM MODEL OF WAGE DETERMINATION IS BARGAINING OVER MATCH SURPLUS and has been for a long time now.

here is an excerpt from these slide
http://economics.mit.edu/files/7400

Can wages be equal to marginal cost of labor and value of marginal product of labor?

No

[Luis Enrique]

also, the claim that marginal productivity is the basic driver of wages does not come with the corollary of perfect information etc. I mean you can think that marginal productivity is the basic principle at work, but also that outcomes are stochastic, there are unobservable quality of matches etc. etc. so I don't think you can start from observing those things and work backwards to the conclusion that marginal productivity is not the basic driver of wages.

I don't want to get into the semantics of what equilibrium means in economics, but I don't think your observation "life is lived in disequilibrium" quite gets at it (if you simulate at DSGE model, it never actually is in equilibrium, it's still an equilibrium model) and I think if you built a model of football still based on idea that clubs will attempt to pay the players their marginal product but add in a few realistic features (such as unobservable complementaries between players, stochastic outcomes) you will end up with a model that exhibits many of the things you say here invalidate marginal product theory, even though the model contains that theory by construction.

[Magnus]

Luis is right.

W = MP in first year undergrad textbook - written and presented with multiple caveats that this really is not reality and should not be interpreted as reality - really does not mean that "mainstream economics assumes that everyone gets paid their marginal product."

As well as the references Luis gives check out

http://piketty.pse.ens.fr/files/PikettySaezStantcheva2014.pdf

published in the American Economic Journal, also about as mainstream as it gets.

Please read and digest the following sentence written in the abstract of this paper: "US CEO “pay for luck” is quantitatively more prevalent when top tax rates are low".

Now ask yourself - is this mainstream publication asserting that W = MP? No it is not. It is saying that wages of CEOS is determined by luck. Read it again if you don't believe.

Yes but this must be some heterodox journal. No it isn't. The AEJ is published by the American Economic Association and is dead centre mainstream.

Sheesh will the blogosphere ever get this point.

[Answer - no: because it's populated by armchair philosophers who in order to propagate their wankery thinking have a vested interest in undermining the idea that resource allocation can ever be analyzed scientifically - which is what mainstream economists, actually, are doing.]

[gastro george]

@Luis - interesting response. I wonder if part of what might be called the "what-is-mainstream" debate is due to people talking past each other in mutual miscomprehension, and this might be an example.

You describe nicely at the end how it's possible to build a model based on MPT that might describe the situation in football. But an anti-mainstreamist might say that this is not how wages are actually set in real life, as it's not possible to calculate marginal values. So, while the model might be accurate, it's not describing what actually happens. So, it might be argued, what is the model contributing?

Just my tuppence. Might be rubbish.

[Luis Enrique]

Hi George,

well let's just say the model is merely a way of saying "this is how we think things work" and contributes no more than that. My guess is that in football, whilst everybody will always try to sign players and pay them as little as they can get away with (i.e. employers like paying less than MP) player bargaining power is generally quite strong, so they will get close to what their *perceived* MP is, and to my mind at least, I think it's quite realistic to suppose that clubs do think things like, if I sign Ronaldo, how much will that add to shirt sales, ticket sales, how much will that alter my chances of winning things etc. and hence they are thinking in marginal revenue terms in real life, even if you can never (of course) calculate it as such. But then, also of course, with football is also a hobby of oligarchs, and the marginal value of a shiny trinket to a rich collector of trinkets is another way of looking at it.

[Maurits Pino]

As a criticism of the neoclassical theory, the football comparison makes little sense, IMO.
W=MP because a price taker in the labour and goods market expands the workforce and production until the condition holds (or reduces it until ...). With team size fixed (at least: the number of players in the field) that doesn't work very well.

[Bob]

IMO the problem is there is no 'no deal' option for the individual.

Businesses only hire if they can make a profit. People need to get hired to eat and live. That's a power imbalance that has to be corrected to have a fair market.

Therefore it isn't a fair market.

9 bones and 10 dogs is completely unfair. The job of the state is to fix that power imbalance by making sure there are enough jobs via a job guarantee. Then you get fair pricing.

Then businesses have to *compete* to get staff. 'No deal' is an option for both private parties. Fair competition is no robbery.

And of course people with more money in their pocket spend more, which increases the turnover and profit of businesses.

[Metatone]

We've had this discussion before, Luis Enrique, but I'll put in my usual caveat:

The important economists are the ones who set policy. And mysteriously, they seem, over the last 30 years or so to have almost always come not out of the 80% sensible ones you mention, but the other 20%.

Here's a nice recent paper on some of the mechanics of this in the UK context:

http://onlinelibrary.wiley.com/doi/10.1111/1467-9248.12198/abstract

[Luis Enrique]

Metatone

I don't really know much about the policy making process, and if I had time would read that paper you link to with great interest.

However I have met a fair number of economists who work for the Treasury, past and present, and academic economists who are involved in policy making, and when it comes to how wages are determined at least, I'd place them in the 80. Here is an example
http://www.bath.ac.uk/sps/staff/paul-gregg/
take a look through his work, he's not carrying around daft ideas of how labour markets work.

[Dennis Smith]

I'm not sure that this answers Metatone's point. No doubt the Treasury employs many sane and competent economists: there are many posts where their skills are required. The key question is whether these are the economists who have the ear of ministers and other decision-makers.

Mrs Thatcher famously used to ask of civil servants: is he one of us? In plain language, she politicised the civil service. What we need here is some assurance that policies are selected for their economic validity, not their ideological consonance.

[chris]

@ Luis, Magnus - I take your point. My 1st par might have been straining for effect. I'm happy to agree that very many proper economists agree that wages are determined by bargaining. My post was not intended as an attack upon "mainstream" economics, but upon those simple-minded souls who try to use marginal product theory to justify high pay. (The fact that one of those souls was the author of a well-known mainstream textbook is a point I'll overlook.)

[Dave Timoney]

The irritation of the economists is misplaced. I suspect Chris is making a joke here at the expense of Mervyn King, a man whose view of the world was heavily influenced by having to deal with idiots like Fred Goodwin and who then found himself faced (albeit briefly) with idiots like Randy Lerner and Joleon "a weight off our shoulders" Lescott.

Despite Chris's claim that "this is not an idiosyncratic example", I think most people would concede that the peculiarities of remuneration in the banking and football industries cannot easily be explained by mainstream economics, though political economy (specifically public choice theory) makes a decent fist.

Chris makes the point about bankers being overpaid to prevent stealing. I'd add that footballers' wages (treated as a package with transfer fees) can be thought of as an indirect form of voting - i.e. an expression of supporter preferences - which is why low-paid fans demand that the board "spend some bleedin' money".

Villa's problem is that the fans think they're being denied this voice, through Lerner's reluctance to spend at what they consider the "right" level for a club of its pedigree (it's wage-bill is 12th out of 20), hence they have headed for the exits (a la Albert Hirschman).

[Peter Dorman]

The claim that most economists who work directly or very closely with wage determination are not believers in simple MP theory is correct. However:

(1) Textbook Econ 101 matters. It is what most people who learn a smattering of economics (which includes most politicians and journalists) learn. It also provides the default conceptual frame for the field.

(2) It is very common in applied work for economists to employ the wage as a proxy for unmeasured productivity, even in labor economics (not to mention growth theory). How should this be understood?

[Egmont Kakarot-Handtke]

Marginalism is the landmark of scientific incompetence
Comment on ‘Limits of marginal productivity theory’

You conclude: “It could be that marginal product theory — just like simple-minded talk of incentives — is as much ideology as science.”

It could also be that it is merely the usual brainless waffling economists’ are widely known for.

Standard economics is built upon this set of hard core propositions, a.k.a. axioms: “HC1 economic agents have preferences over outcomes; HC2 agents individually optimize subject to constraints; HC3 agent choice is manifest in interrelated markets; HC4 agents have full relevant knowledge; HC5 observable outcomes are coordinated, and must be discussed with reference to equilibrium states.” (Weintraub, 1985, p. 147)

Marginalism follows logically from the green-cheese behavioral assumption of constrained optimization HC2. What can be said with certainty is that the whole set of Walrasian axioms is methodologically inadmissible. By implication, standard marginalistic distribution theory falls flat.

It is of utmost importance to realize that the concept of marginal productivity is dead. So it is a welfare-diminishing waste of time to mention, to criticize, and to discuss it.

The root defect of the familiar distribution theories is that the representative economist cannot even tell the difference between income and profit (2012; 2014). This is not exactly a noteworthy achievement.

Marginal distribution theory is the widely visible landmark of economists’ scientific incompetence. The elimination of the misery demands a paradigm shift from microfoundations as embodied in HC1/HC5 to macrofoundations.

Egmont Kakarot-Handtke

References
Kakarot-Handtke, E. (2012). Income Distribution, Profit, and Real Shares. SSRN Working Paper Series, 2012793: 1–13. URL http://ssrn.com/abstract=2012793.
Kakarot-Handtke, E. (2014). The Profit Theory is False Since Adam Smith. What About the True Distribution Theory? SSRN Working Paper Series, 2511741: 1–23. URL http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2511741.
Weintraub, E. R. (1985). Joan Robinson’s Critique of Equilibrium: An Appraisal. American Economic Review, Papers and Proceedings, 75(2): 146–149. URL
http://www.jstor.org/stable/1805586.

[Deviation From The Mean]

Aside from the more concrete laws of wages, the structure of a capitalist firm requires a certain stratification. This stratification did not develop as a result of marginal product but as a result of the need to discipline workers - like the hierarchy of an army is based on necessity and not laws of marginal efficiency. A by product of this history is managerialism and the great leader so beloved of the neo-liberals.

The reality is that the pay of top executives exploded while real wages for workers stagnated. But these periods are inevitable under conditions of capitalism and competition.

It doesn't really matter whether mainstream economics believes theory x or y, the fact is it denies the exploitative nature of capitalism and comes up with all sorts of extravagant mathematical meta theories to say capitalism is anything but exploitation.

So mainstream economics does indeed suck.

Luis, being a sycophant of the mainstream, i.e. the ruling class, builds a straw man in his defence of the mainstream and ignores the real vulgar nature of the mainstream.

[Blissex]

«THE MAINSTREAM MODEL OF WAGE DETERMINATION IS BARGAINING OVER MATCH SURPLUS and has been for a long time now.»

Then either the Lucas Critique is a stupid joke or the models taught and used by the mainstream are microfounded on «WAGE DETERMINATION IS BARGAINING OVER MATCH SURPLUS» and not over a single agent trading with itself in infinite markets and paying itself wages according to its marginal productivity.

If so I am amazed but relieved to learn that Arrow-Debreu(-Lucas) [or similar] models are no longer taught in textbooks and universities as the apex of eternal Economics, and DSGE models based on Arrow-Debreu(-Lucas) [or similar] are not used by any major central bank, treasury, international organization to form policy recommendations. And when papers based on Arrow-Debreu(-Lucas) [or similar] models are submitted to major journals they get instantly binned because everybody knows that «WAGE DETERMINATION IS BARGAINING OVER MATCH SURPLUS», and when famous professors go on about marginal productivity as determining compensation or using it in their models "the audience starts to whisper and giggle to one another".

Or else there is much buffoonery around, and papers about «BARGAINING OVER MATCH SURPLUS» happen only as self-contained incidents like the Cambridges Capital Controversy was regarded by most Economists, and the cardinal verities are unchangeable.

[Blissex]

«The claim that most economists who work directly or very closely with wage determination are not believers in simple MP theory is correct.»

Indeed, and as long as they use the cover of jargon and technicalities and don't explicitly connect the dots and undermine the central truthinesses of Economics it is indeed possible for specialists in one narrow field of Economics to publish "heresy" even in mainstream journals.

As to that I think that there are "gold standard" published papers that contradict essentially all the assumptions and conclusions of the dominant mainstream model(s), but that has not stopped many from teaching and using those models.

And that's perhaps thanks to dear old bulls-hitter M Friedman's "Methodology of Positive Economics", as to which J Tobin wrote "has done great damage".

[Luis Enrique]

Chris, understood ta. Yes the less said about him the better.

Dennis, agreed.

Dftm, a model in which bosses have all the bargaining power is a model of exploitation

Blissex I don't follow the inconsistency between wage determination by bargaining and the Lucas critique, there are plenty of DSGE models that incorporate such a form of labour market. There *was* a great deal of sniggering and derision when famous professor who is not being named came out with his MP based defence of inequality, however the idea that productivity may explain a large share of the distribution of wages, is a different, more sensible, proposition. For example:
http://economics.mit.edu/files/7490

[An Alien Visitor]

The dominant mainstream model(s) have had huge amounts of resources thrown at them, no surprise as the dominant mainstream model(s) all justify and provide ideological cover for the existing class relations and the power of capital over labour.

Economics is strange in that it is a scientific subject in the sense of methodology, published papers, journals, peer review etc but in terms of content economics is purely ideological. and stays within the narrow bounds of class interests, and are imbued with class prejudice.

Those who believe science is method really need to take a close look at economics.

[Blissex]

«Textbook Econ 101 matters. [ ... ] It also provides the default conceptual frame for the field.»
«It is very common in applied work for economists to employ the wage as a proxy for unmeasured productivity,»

This is sort of a argument that mainstream Economics is a gnostic religion, with a public doctrine for the masses, which is not merely approximate but quite wrong (a "noble lie"), and a different doctrine for the initiates, that gets taught only in postgraduate seminars or in specialist journals, and that microfounded macroeconomists are not initiates and ingenuously base their models on the quite wrong public doctrine while the initiates of course keep very quiet about that.

[Deviation From The Mean]

"a model in which bosses have all the bargaining power is a model of exploitation"

Please, do me a favour. It is a model which assumes wage slavery is a permanent condition of existence - i.e. exploitation is a given.

It is purely in the service of exploitation. It recognises that over exploitation undermines exploitation, without recognising that periods of over exploitation are an essential condition of an exploitative system.

[Blissex]

«there are plenty of DSGE models that incorporate such a form of labour market.»

If I understand correctly, you are claiming that "mainstream" DSGE models are as rule not models of general equilibrium, at at least at not microfounded on Arrow-Debreu(-Lucas) [or similar] models.

I am greatly relieved by these news :-).

[Blissex]

«don't follow the inconsistency between wage determination by bargaining and the Lucas critique»

My simplistic understanding of the Lucas Critique is that ad-hoc, evidentially based models are not intellectually respectable for Lucas and the "mainstream", and models of Economics must therefore be microfounded on "first principles" axioms; and that the only "mainstream" acceptable set of such axioms are those from which the "mainstream" derives Arrow-Debreu(-Lucas) [or similar].

Then how is «bargaining over match surplus» a valid "mainstream" microfoundation?

Perhaps a set of "from first principles" axiom can be found under which «bargaining over match surplus» can be derived, and thus be a valid but non-"mainstream" microfoundation, or we can abandon the strictures of the Lucas Critique and say that the evidence shows that «bargaining over match surplus» just happens in practice.

[Luis Enrique]

Blissex,
well I have always understood the Lucas critique to mean that it's a bad idea to base predictions of how things will respond to a policy change on the basis of empirical regularities in historical data because when new policies are introduced, behaviour may change and the relationships between variables with it. Whereas if you have a model of how people behave and why, it can - in theory - explain how things turn out under one policy and how people'd behave and things would turn out in another. If so, if you think you microfounded actors face 'frictions' or other reasons why W=MP doesn't work, then a model of how people behave and why could be written that includes that, and avoid the Lucas critique. But I write this tentatively, could well be wrong. Also I think you can still call a model "general equilibrium" because it tries to capture all effects on everything once all follow-on effects have followed-on (as opposed to partial equilibrium which holds stuff constant) even if it includes various features that mean markets don't clear. that is to say, I don't think general equilibrium means all markets clear. But don't quote me on that.

[Blissex]

«productivity may explain a large share of the distribution of wages, is a different, more sensible, proposition.»

That's an ad-hoc, evidentially based story, and it has nothing to do with the claim that "mainstream" Economics makes that the (Pareto-optimal!) distribution of income (among the sole agent and itself!) is determined solely by productivity (and initial endowments)

The "mainstream" as I understand it does agree with the contrary happenstance; indeed nobody here (I hope) is denying that the "mainstream" teaches that it so happens in practice that (because of distorting government oppression) *in practice* the distribution of income only *mostly* reflects productivity.

I am sorry if there has been a misunderstanding about that.

[sam]

Luis Enrique speaks the truth, as usual.

Lars Syll is wrong, we can observe wages marginal product of labor. Or at least one clever PhD student figured out a convincing way. Turns out wages are about 85% of marginal product. https://bepp.wharton.upenn.edu/bepp/assets/File/Isen%20jmp.pdf

[Blissex]

"anonymous" :-) opined:

«don't think general equilibrium means all markets clear»

Whatever :-). But we are discussing a narrower topic, what matters here are "mainstream" models specifically.

I think we can agree that "mainstream" means "general equilibrium" (or else this is not a productive discussion for me), then I claim that the relevant notion of "general equilibrium" is what it means for the "mainstream", not for you or me or someone else.

And even more particularly whether there is a specific approach to "general equilibrium" models that is "mainstream", which I think is Arrow-Debreu(-Lucas) [or similar] and whether «bargaining over match surplus» is compatible with that, instead of being merely one of the many special-case happenstances that like so many others can get published in "mainstream" journals as long as it is protected by jargon and absence of connection to what matters most. What is "mainstream" for labor Economics is not necessarily "mainstream", and can well be incompatible with it. Compartmentalization can be a source of much productivity :-).

[Blissex]

Another line that may cut pretty close to the core of the matter: the bulls-hitting central theorem of "welfare" Economics, the one that allows "mainstream" Economists to consider only "efficiency" and disregard "distributional impact".

Isn't that "mainstream"? If it is not "mainstream", then what is the "mainstream"?

And isn't it critically based on Arrow-Debreu(-Lucas) [or similar] axioms? As to that I may misremember about the technicalities, but I guess that there is a big chance that the central theorem of "welfare" Economics only holds in models where marginal productivity (and initial endowments, which are rarely mentioned) determines distribution of income (absent distorting government oppression).

And if the central theorem of "welfare" Economics does not hold, what happens to "mainstream" theorizing? Isn't it wholly to rebuild?

[Blissex]

«one clever PhD student figured out a convincing way. Turns out wages are about 85% of marginal product.»

Interesting paper on a "natural experiment".

My quick take is that the author figured out that in the case of workers subject to close monitoring in a small firm, managers/owners in general know pretty well how much they can get out of each employee in terms of "revenue", and use that information to systematically underpay them. This is little more than the statement that each manager/owner has a budgeting spreadsheet with salary costs, target revenues, and a target for profitability, and adjusts pay levels to hit that target. Note also the appendix data, as the author says "analysis is limited to small firms":

Average number of employees: 7.33 (8.14)
Average annual payroll: $281,245 (541,807)
Average annual revenue: $1,312,767 (9,500,126)

That is a broad spectrum, but $38,000 average wage.
Also note that payroll is on average just 21% of revenue. Uhm... There is still quite a wide variation in degree of underpayment.

But from a theoretical point of view most importantly treating "revenue" as if it were "productivity" begs the whole question.

Also my prejudice is that papers that talk about "Cobb-Douglas" are dislikable (question begging on a deep bender), even if the paper claims "the results do not rely on a production function with that specific functional form".

[Kevin Carson]

Clark's "marginal productivity" theory is circular -- a factor input's "marginal productivity," by definition, is whatever it adds to the final price of the finished good. Which means that marginal productivity is determined by what the factor owner is in a position to charge.

And in the case of labor, figuring backwards from marginal productivity to input price isn't just circular -- it's absolutely backwards. Labor is unique in that its input cost is not determined straightforwardly like that of inanimate material inputs, but by the subjective disutility to the laborer of providing it. At least that would be the case in a free market.

When the laboring classes historically have been systematically expropriated and evicted from the land, and the state acts in collusion with capital to create a system of power that reduces the bargaining power of labor and create a buyers' market for labor, the pricing of labor shifts to a greater or lesser degree from disutility to the subsistence cost of reproducing labor-power.

But in a situation where labor remained in possession of the land and capitalist employers had to seriously compete with the possibility of self-employment, the "marginal productivity" of labor would be determined by whatever the necessity of making it worth their while to work added to the price of finished goods.

[Keith]

There is an oddity about economic theory. In Physics there is a simple general rule. Variations based on specifics can be added but do not invalidate the rule e.g. Gravity.

In economics the rule has twenty exceptions and it turns out most observed cases are covered by exceptions. So the rule is unprovable or only works in a small number of cases. Is this a theory? Can we take it seriously if it seldom maps to observation? Does saying expected revenue determines Footballer pay make sense?

[reason]

Isn't there a more basic problem here. It is not YOUR marginal product that you paid, it is the marginal product of any potential replacement. They don't have to offer you anymore than that, because nobody other than them knows how much you are worth.

[Luis Enrique]

Keith,

I don't think it's an "exception" to say that real life contains, for example, errors, so even if your theory is y = f(x) what you observe is y = f(x)+e

it's not the fault of economists that the world is complicated and no theory is going to completely explain all we observe. that is true of the theories that Kevin or Dftm, above, believe too.

It is hard to distinguish between competing theories, I couldn't say exactly what level of empirical support is required to take a theory seriously, it depends what you mean by "seldom maps to observation" - theory will *never* may to observation precisely, so it's always a question of how much error you tolerate.

reason, I think that's why you need competitive market, so the MP of your potential replacement equals your MP. If you have in mind a model in which workers are not perfect substitutes, productivity is match specific, then you are in a world of bargaining over match surplus. That doesn't mean productivity is irrelevant, but it becomes something that set bounds on feasible wages, not something that pins it down.

[Luke]

Don't efficiency wages apply to footballers? You have to pay them loads to prevent them throwing matches.

[Egmont Kakarot-Handtke]

The zero productivity of economists

All feathers are subject to the Law of Gravitation, but the trajectory of a flying feather on a windy day is a random walk with a downward bias. The underlying law is totally obscured by historical accident. The very characteristic of a scientist is to abstract from unique historical accident while the dilettante is inextricably glued to it. Every economist could know since from J. S. Mill.

“Since, therefore, it is vain to hope that truth can be arrived at, either in Political Economy or in any other department of the social science, while we look at the facts in the concrete, clothed in all the complexity with which nature has surrounded them, and endeavour to elicit a general law by a process of induction from a comparison of details; there remains no other method than the à priori one, or that of ‘abstract speculation.’” (1874, V.55)

To tackle the problem of distribution by looking at a concrete case of wage setting at Aston Villa is a fine example for the methodological blunder that is rampant in economics. This nuisance comes under the general heading of methodological individualism.

“It is a touchstone of accepted economics that all explanations must run in terms of the actions and reactions of individuals. Our behavior in judging economic research, in peer review of papers and research, and in promotions, includes the criterion that in principle the behavior we explain and the policies we propose are explicable in terms of individuals, not of other social categories.” (Arrow, 1994, p. 1)

This research program has abysmally failed. The scientific productivity of economists has been consistently zero since Jevons/Walras/Menger. How does it come that their wages have been greater zero? In economics, to be sure, there is no relationship between performance and reward at all.

Economics is not a science but an employment program for wannabe scientists. Keynes once proposed “to fill old bottles with banknotes, bury them at suitable depths in disused coalmines” and then “to dig the notes up again”. This is pretty much the same thing as discussing the marginal productivity theory over and over again. To make economics a science first of all requires to bury economists “at suitable depths in disused coalmines.”

Methodological individualism is finished. Marginalism is finished. For the correct distribution theory see (2014).

Egmont Kakarot-Handtke

References
Arrow, K. J. (1994). Methodological Individualism and Social Knowledge. American Economic Review, Papers and Proceedings, 84(2): 1–9. URL http://www.jstor.org/stable/2117792
Kakarot-Handtke, E. (2014). The Profit Theory is False Since Adam Smith. What About the True Distribution Theory? SSRN Working Paper Series, 2511741: 1–23. URL http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2511741
Mill, J. S. (1874). Essays on Some Unsettled Questions of Political Economy. On the Definition of Political Economy; and on the Method of Investigation Proper To It. Library of Economics and Liberty. URL http://www.econlib.org/library/
Mill/mlUQP5.html#EssayV.OntheDefinitionofPoliticalEconomy.

[Nanikore]

@Luis you say,

"and that if markets competitive they will get what they are worth .. but then labour markets are very far from competitive"

That's not the problem. In many ways labour markets, with the removal of many border controls and the virtual elimination of effective trade unions, are more flexible and competitive than ever. We have labour mobility on a scale we have never seen before. The EU since its expansion is a prime example. In fact this is largely what brought on the Brexit referendum.

But if anything prices have even less relationship than with marginal cost.

We need to get away from this framework as being the basis of analysis and then arbitrarily adding frictions. There is no reason why markets SHOULD work on the basis of marginal productivity theory. We need proper ontological and epistemological justification rather than imposing this framework on all analysis. Once we can get rid of this religious clutter, we can start analysing what actually goes on.

[Blissex]

"Nanikore" and others, the discussion here is whether the marginal productivity *theory* of income distribution is *currently* part of *mainstream* Economics.

There is no doubt that mainstream Economists today concede that often it does not hold in practice even if it were true in theory because of distorting government intervention (for example "heroes of wealth creation" are paid well below their marginal producitvity), and that in the past it was part of "mainstream" Economics.

Whether the theory of distribution of income based on marginal theory actually is true, even approximately, is a completely different topic.

"Luis Enrique" seems to me mainly to claim that such theory is not part of "mainstream" Economics (note: not of mainstream labor Economics) today, and it has been abandoned, and replaced with a completely different theory, so it is pointless to argue as to its merit.

His secondary point seems to me about the paper mostly looking at low wage workers in small shops is neither here not there as to that, or even as to the actual validity of marginal productivity theory whether in theory or in practice.

[Geoff Willis]

Top pay scales, like all income distributions are power law distributed. This fact means that they simply can't be distributed according to marginality.

See Gabaix:

http://www.nber.org/papers/w14299

and Willis:

http://www.tandfonline.com/doi/abs/10.1080/02692171.2015.1065225#.Vxh2B-YdzIU

[Luis Enrique]

"There is no doubt that mainstream Economists today concede that often it does not hold in practice even if it were true in theory because of distorting government intervention"

no, I do not think "distorting government intervention" is why most mainstream economists think W=MP does not hold in practice. I think most mainstream economists would talk about rents extraction, bargaining power, market failures etc. And most would say CEOs etc. are paid *above* their MP (or actually, probably that MP is the wrong way of thinking about CEOs).

the link between productivity and wages has not be abandoned, but changed. I talked about bargaining over match surplus. Match surplus is productivity.

I didn't link to a paper about small shops.

Geoff, Gabaix deserves to be far more famous imo and your ABM work looks v interesting.