How Economics Sees Reality

Something has been bothering me about the way evidence is (sometimes) used in economics and econometrics:  theories are assumed throughout interpretation of the data. The result is that it’s hard to end up questioning the model being used.

Let me give some examples. The delightful fellas at econjobrumours once disputed my argument that supply curves are flat or slope downward by noting that, yes, Virginia, in conditions where firms have market power (high demand, drought pricing) prices tend to go up. Apparently this “simple, empirical point” suffices to refute the idea that supply curves do anything but slope upward. But this is not true. After all, “supply curves slope downward/upward/wiggle around all over the place” is not an empirical statement. It is an interpretation of empirical evidence that also hinges on the relevance of the theoretical concept of the supply curve itself. In fact, the evidence, taken as whole, actually suggests that the demand-supply framework is at best incomplete.

This is because we have two major pieces of evidence on this matter: higher demand/more market power increases price, and firms face constant or increasing returns to scale. These are contradictory when interpreted within the demand-supply framework, as they imply that the supply curve slopes in different directions. However, if we used a different model – say, added a third term for ‘market power’, or a Kaleckian cost plus model, where the mark up was a function of the “degree of monopoly”, that would no longer be the case. The rising supply curve rests on the idea that increasing prices reflect increasing costs, and therefore cannot incorporate these possibilities.

Similarly, many empirical econometric papers use the neoclassical production function, (recent one here) which states that output is derived from the labour and capital, plus a few parameters attached to the variables, as a way to interpret the data. However, this again requires that we assume capital and labour, and the parameters attached to them, are meaningful, and that the data reflect their properties rather than something else. For example, the volume of labour employed moving a certain way only implies something about the ‘elasticity of substitution’ (the rate at which firms substitute between labour and capital) if you assume that there is an elasticity of substitution. However, the real-world ‘lumpiness‘ of production may mean this is not the case, at least not in the smooth, differentiable way assumed by neoclassical theory.

Assuming such concepts when looking at data means that economics can become a game of ‘label the residual‘, despite the various problems associated with the variables, concepts and parameters used. Indeed, Anwar Shaikh once pointed out that the seeming consistency between the Cobb-Douglas production function and the data was essentially tautological, and so using the function to interpret any data, even the word “humbug” on a graph, would seem to confirm the propositions of the theory, simply because they follow directly from the way it is set up.

Joan Robinson made this basic point, albeit more strongly, concerning utility functions: we assume people are optimising utility, then fit whatever behaviour we observe into said utility function. In other words, we risk making the entire exercise “impregnably circular” (unless we extract some falsifiable propositions from it, that is). Frances Wooley’s admittedly self-indulgent playing around with utility functions and the concept of paternalism seems to demonstrate this point nicely.

Now, this problem is, to a certain extent, observed in all sciences – we must assume ‘mass’ is a meaningful concept to use Newton’s Laws, and so forth. However, in economics, properties are much harder to pin down, and so it seems to me that we must be more careful when making statements about them. Plus, in the murky world of statistics, we can lose sight of the fact that we are merely making tautological statements or running into problems of causality.

The economist might now ask how we would even begin to interpret the medley of data at our disposal without theory. Well, to make another tired science analogy, the advancement of science has often not resulted from superior ‘predictions’, but on identifying a closer representation of how the world works: the go-to example of this is Ptolemy, which made superior predictions to its rival but was still wrong. My answer is therefore the same as it has always been: economists need to make better use of case studies and experiments. If we find out what’s actually going on underneath the data, we can use this to establish causal connections before interpreting it. This way, we can avoid problems of circularity, tautologies, and of trapping ourselves within a particular model.

About these ads

, , , ,

  1. #1 by Rutger on October 13, 2013 - 1:26 pm

    Unlearningecon wrote:
    “economists need to make better use of case studies and experiments. If we find out what’s actually going on underneath the data, we can use this to establish causal connections before interpreting it. This way, we can avoid problems of circularity, tautologies, and of trapping ourselves within a particular model.”

    But then they would have to give up their physics-envy, and admit that the economy is embedded in social relations (and thus subject to change) and not the ultimate function of timeless and unchanging laws of supply and demand?

  2. #2 by Cameron Murray (@Rumplestatskin) on October 13, 2013 - 2:04 pm

    I just wrote this

    http://ckmurray.blogspot.com.au/2013/10/economists-do-it-with-models-is-one-of.html

    checked twitter, and came here. Very much on the same wavelength. Eerie.

    Good work again.

    • #3 by Unlearningecon on October 13, 2013 - 3:13 pm

      Haha, excellent!

    • #4 by gastro george on October 13, 2013 - 4:21 pm

      “No matter how much your model appeals to your intuitive reasoning, or how well it fits the data, it cannot be shown to be of scientific value unless it offers useful predictions.”

      … should be tattooed on a few foreheads.

      • #5 by BRAGIN, JOHN on October 15, 2013 - 12:49 am

        Depends on what you mean by ‘predictions’. If you mean medium and long range exact, point predictions, then we are out of luck, these are simply not possible in non-linear systems. And when probabilities are involved not even short term point-predictions are possible. That is a proven fact. If you mean projections of trends or aggregates then we know that we have tools that can tell us under what conditions our models can give us reasonable predictions of aggregates or trends to what time horizons into the future. To demand of our theories and models some traditional old-fashioned LaPlacian ideal in social science is even to misunderstand LaPlace’s writings. I take the position of Joshua Epstein in complex systems science, the generative scientist’s motto: “If you can’t grow it, you can’t understand it.” Where explanation is more basic than prediction (in any sense), and where we must strive for what prediction we can get.

  3. #6 by BFWR on October 13, 2013 - 6:15 pm

    Economics needs to go deeper and more basic for its understandings. What can be more basic in a monetary economy than cost and price, (money required) and individual income and price? (exchange) These are the physics and most deeply embedded basics of economics. Decipher the enforcements and relationships there….and everything else is the chaos of human action that is up for conjecture.

    The real problems with economics is that economists are able to get their degrees without so much as a basic accounting course. Thus they have lost track of the basics, the empirical data of cost/price and its relationship to individual income. Economists like Steve Keen have recently become aware of the value of accounting/double entry bookkeeping….in general. Now if he and other economists will look a little bit closer and decipher the data and the economic relationships to be found in COST accounting…they’ll be looking at the most elemental and unchanging realities of economics/the productive process. So be it!

    • #7 by Cameron Murray (@Rumplestatskin) on October 13, 2013 - 10:56 pm

      “The real problems with economics is that economists are able to get their degrees without so much as a basic accounting course. Thus they have lost track of the basics, the empirical data of cost/price and its relationship to individual income. ”

      Yep. The best economists (in my view) tend to have broad life experience in business or government, and a decent familiarity of law and accounting. Before getting into economics proper I had part of an engineering degree, and a degree in property economics, which included study of property valuation methods, marketing, accounting, financial analysis, and law. At the time it felt little weird being such a generalist, but looking back now EVERY economist needs some grounding in these basics of business and commerce.

      This lack of connection and consistency with other business (and social science) disciplines jumps right out at you.

      http://ckmurray.blogspot.com.au/2013/10/top-young-economists.html

  4. #8 by notsneaky on October 13, 2013 - 9:57 pm

    Anwar Shaikh once pointed out that the seeming consistency between the Cobb-Douglas production function and the data was essentially tautological, and so using the function to interpret any data, even the word “humbug” on a graph, would seem to confirm the propositions of the theory, simply because they follow directly from the way it is set up.

    Ugh, not this again. The word “tautological” means something else than you think it means.

    http://digamo.free.fr/solow74.pdf

    • #9 by Unlearningecon on October 13, 2013 - 10:55 pm

      OK, that’s a fine rebuttal. However I think my overall point still stands and Solow (as well as other economists) have actually expressed similar reservations about the use of the PF.

      • #10 by notsneaky on October 14, 2013 - 12:11 am

        Cameron, I wasn’t disputing that statement by UE. I was disputing the statement of UE which I actually quoted.

        As to the statement you quote, well, that’s where the first sentence of that next-to-last paragraph sweeps the main issue under the rug.

    • #11 by Cameron Murray (@Rumplestatskin) on October 13, 2013 - 11:01 pm

      I actually think what UE wrote about this is correct and consistent with what Solow wrote in your link. Solow says

      “It merely shows how one goes about interpreting given time series if one starts by assuming that they were generated from a production function ”

      While UE says

      “However, this again requires that we assume capital and labour, and the parameters attached to them, are meaningful, and that the data reflect their properties rather than something else.”

  5. #12 by First Financial Insights on October 13, 2013 - 10:04 pm

    Economics simply fails because it ignores the laws of physics, exponential mathematics and the concept of eternity. It falsely believes its abstractions and theories, which create a make-believe positive-sum-game world, will, by sheer hope and dreams, overcome and defeat the negative-sum game constraints that define existential reality. It AIN’T going to happen – Entropy and Eternity will win the day – as absolutes you cannot convince, negotiate or legislate in any way.

    When Einstein said he was not sure about eternity, but certain there was no end to human stupidity – there is little doubt he was thinking specifically about economics with its beliefs, gurus and sheep…

    So the writing is on the wall, as economies all over the world are beginning to look more like Nauru or Easter Island with every passing day. From Cyprus and Iceland, to PIIGS, to Egypt and Syria and now to even Venezuela or possibly Japan – what is perfectly clear is that when you run out of inputs (i.e. resources) it is utterly impossible to produce outputs, such as food, regardless of how much Keynesian currency and debt you print and create, to feed your hungry civilians. “Again, it just AIN’T going to happen.”

    Now you try explaining that to an Economist

    T A McNeil
    First Financial Insights Inc.

    • #13 by A on October 16, 2013 - 7:12 pm

      How did you get to know that Einstein´s statement was speaking of/about economics? Thanks.

      • #14 by First Financial Insights on October 19, 2013 - 4:14 am

        Whom else or what other group would come first to mind in the thoughts of a physicist? Try doing a straw poll yourself and see what happens. The ones, and many other scientists, including top nuclear physicists, that I’ve had the pleasure of dealing with, have quietly confirmed that economists carry an unqualified insanity diagnosis because they ignore or do not understand the absolute realities, not only of entropy; but also, exponential mathematics. They scratch their heads in utter disbelief. “they can’t even apply and integrate such simple concepts to their macro doctrines” True stories – they really think every single Economists is just Nuts – or uses cheap drugs!.

        Personally, I do not understand why they think like that and any clarification would be appreciated. TM

      • #15 by A on October 19, 2013 - 5:54 am

        I really have no clue as to what other group would come first to mind in the thought of a physicist…i´m not even sure whether a physicist could have a hint concerning the deeply flawed state-of-art within Economics…

        Being an economist myself, and having spent a rather long time studying math as an autodidact, it seems to me quite clear that an improved math-understanding would enable a different path in the profession…

        The following question could yield a simple rule upon which a economist´s dishonour might be openly claimed: “Are you by any chance at all aware of the sort of math-foundations of, let´s say, practically rival theories such as microeconomics, on one hand, and macroeconomics?”

  6. #16 by notsneaky on October 14, 2013 - 1:31 am

    Btw, while we’re on the subject of funny production functions, then it’s worth noting that pretty much all the aggregation-related criticisms that get thrown at “mainstream economics” (Cambridge, SMD theorem) don’t go away if one takes a different approach (like, say, Sraffa). If you try to have your theory as general as possible, you get “general” results. Like the “Joan” production function:

    http://www.econ.upf.edu/~mcolell/research/art_065.pdf

    • #17 by Unlearningecon on October 14, 2013 - 10:37 pm

      You’re right, capital is an incredibly tricky concept and I don’t think anybody (bar perhaps Marx) has managed to figure it out theoretically.

      Having said that, I regard Sraffa’s work essentially as a critique of neoclassicism, so he assumed some of the same mechanics (such as constant technology) in order to show what he wanted to show.

      • #18 by notsneaky on October 15, 2013 - 12:50 am

        It’s not so much capital (nm Marx), it’s more about intertemporal processes and aggregation/heterogeneity. I don’t think constant technology has anything to do with it.

      • #19 by Unlearningecon on October 20, 2013 - 1:38 pm

        In the sense that Sraffa’s method of expressing capital as dated labour has to hold technology/productivity constant to be valid then yes, I think technology is one of the problems with defining and aggregating capital.

  7. #20 by Luis Enrique on October 14, 2013 - 12:53 pm

    “This is because we have two major pieces of evidence on this matter: higher demand/more market power increases price, and firms face constant or increasing returns to scale. These are contradictory when interpreted within the demand-supply framework, as they imply that the supply curve slopes in different directions.”

    this is not my field, but from what little I know your claim is false.

    Standard econ model of monopolistic competition and increasing returns to scale (Dixit Stiglitz) predicts, for example, increased demand leads to high production and lower price. [I think. Happy to be corrected]

    separately, I’d like you you explain what market power means in your usage. Power to do what? In standard econ market power (or monopoly power) has to do with your power to raise prices without seeing demand collapse and is measured by inverse of PED. This might have something to do with market concentration (because market concentration might explain your PED) but for example, you could be in a highly concentrated market with say only 2 firms, but if you are selling perfect substitutes and cannot raise prices without demand evaporating, then you don’t have monopoly power, regardless of market concentration. So when you talk about adding a third term, what does it mean? In your world in which mainstream econ is useless, what sort of power does market concentration give firms?

    • #21 by Unlearningecon on October 14, 2013 - 11:09 pm

      Standard econ model of monopolistic competition and increasing returns to scale (Dixit Stiglitz) predicts, for example, increased demand leads to high production and lower price. [I think. Happy to be corrected]

      Sure. But that’s not standard demand-supply.

      no it cannot. For instance it cannot rationalize a decision that raises growth and wipes out profits.

      Sure it can. “They are expanding now in the hope of realising future long term profits”.

      OK so let’s bring the whole PED thing to this thread as the post was partially inspired by that.

      I’d like you you explain what market power means in your usage. Power to do what?

      I’d say: ability to control quantity, price and possibly quality while maintaining market share/profit margin/whatever the firm is aiming for.

      In standard econ market power (or monopoly power) has to do with your power to raise prices without seeing demand collapse and is measured by inverse of PED. This might have something to do with market concentration (because market concentration might explain your PED)

      Yes, various institutional factors might be partially represented by a PED. However, firms don’t actually try to consider or estimate the PED because they don’t consider it an important or feasible concept. Instead, they use things like industry concentration, consideration of advertisement, even what the unions say. Looking at these things independently gives us a far richer view of the pricing process and firm structure than something like the PED. I’m looking to move the theory forward.

      I know you won’t be convinced by anything like this, because no matter what businessmen do they’re really just conforming to the PED analysis. What was that Keynes said about rebuking non-euclidean lines for not keeping straight?

      let’s recap

      Nah, let me do a better recap:

      me: firms don’t use PED or any neoclassical concepts in their estimation
      you: maybe, but the theory still predicts the behaviour of markets right *offers vague data with about a 20% margin of error*
      me: your example shows how tricky precise prediction is in economics. Maybe we should be concerned with actual mechanics and behaviour?
      you: blah blah look you’ve contradicted yourself in some convoluted way haha how would you do this other than with mainstream models etc.

      • #22 by notsneaky on October 15, 2013 - 12:47 am

        However, firms don’t actually try to consider or estimate the PED because they don’t consider it an important or feasible concept.Instead, they use [link here] things…

        firms don’t use PED or any neoclassical concepts in their estimation

        I don’t know how you managed to convince yourself of that but, at least generally, it’s not true (I’m sure you can find some counter examples – I’ll give you one at the end).

        First, the link you provide does not actually show that “firms don’t estimate PED”. It’s just an overview of existing pricing theories.

        Second, here’s a single data point, but I don’t think it’s unusual, and then a follow up question:

        Some years ago I attended a presentation at a university which shall remain nameless given by a head hunter from a large corporation which shall remain.. what the hey, I’ll name it, it was Caterpillar the manufacturer of construction equipment. The guy was trying to recruit econ majors (I was not one of them, went there for unrelated reasons) to apply to the company’s Pricing Division. He stressed the importance of analytic and quantitative skills and when he outlined the responsibilities that new hires would have he… explicitly said “estimating the price elasticity of demand for our products”. In fact I very distinctly remember him saying something like “We are looking for people good in micro”, by which he of course meant neoclassical micro. He also talked about estimating cost curves.

        So there ya go.

        Additionally, I’ve worked for some governmental organizations where the pricing of products of particular industries was key. Guess what. We estimated price elasticities of demand (another division looked at the cost curves). You can click around the publications issued by, say, the US Department of Transportation (not who I worked for) and find lots and lots of applied paper which talk about PED.

        Both public sector and industry do in fact use neoclassical economics, *explicitly* (often of the very crude partial equilibrium Marshallian type). Now, it may very well be true that at the end of the day some manager decides to ignore these estimates and “goes with his gut” or whatever, but the process of actually applying neoclassical economics to pricing decisions is there.

        So here is the follow up question. Suppose a company is hiring for a job which involves figuring out how it should set its prices. One candidate comes in and says “I’m an econ major, I’ve taken Price Theory, Managerial Economics, Industrial Organization. I know about price elasticities, marginal cost curves and models of both perfect and imperfect competition from a neoclassical perspective”. The other guy comes in an says “I’m a heterodox Post-Keynesian economist. I’ve read Kalecki’s later work, and dismiss his earlier work. I’ve read Sraffa, Downward, Andrews and Earl (and whoever else that article you linked to mentions). I believe firms set their prices based on “institutional features””.

        Who do you think they’d hire? Honestly.

        Ok. Third. Somewhere else on this blog, or maybe on your twitter you quoted some body who apparently was at a business school who supposedly laughed at the idea that “marginalism” is taught to MBAs because that would be “ridiculous”. When I saw that I just sighed but here it’s relevant again. Of course MBAs get taught marginalism. Of course they get taught PEDs. Supply and demand. MC=MR. All that stuff. I don’t know who your source was but the Columbia Business School disagrees:

        http://www8.gsb.columbia.edu/courses/mba/2013/spring/b6006-001

        I just picked that one because it was the first hit on google but 99% of MBA programs will have classes on Managerial Economics, and “marginalist” subjects.

        It’s good to approach the world in the critical way. But you really need to think about some of these voluntary blinders you’ve put on yourself which lead you to continually and obstinately saying stuff which is just plain wrong.

        Ok, here’s one famous counter example of a situation where a firm didn’t price in a “neoclassical” way (or did they?). The release of the Nintendo Wii. Priced well below where MC=MR and the markup was (and may even still be) much smaller than suggested by the inverse elasticity (which was very inelastic) rule. Why?

        Here’s another actually. Supposedly Wal-Mart sells its bananas below “cost”, below what they pay their suppliers for them. And they still make a profit on those bananas. How? It has nothing to do with getting you into the store or any kind of cross-complimentarity between goods.

        So yes you can find probably a not insignificant number of situations where pricing decisions are more complicated than simple neoclassical theory would suggest. That’s why we have the field called IO. But in all these cases if you want to understand what is going on you start with the simple neoclassical model (with market power) and extend it in some way.

      • #23 by Luis Enrique on October 15, 2013 - 10:52 am

        “Sure. But that’s not standard demand-supply”

        yes after writing that comment I figured out what you meant: standard demand-supply meaning under perfect competition. And you suggest adding market power would resolve the contradiction. Well yes, as would moving to a model of monopolistic competition with increasing returns. .

        “Sure it can. “They are expanding now in the hope of realising future long term profits”.

        I should have been more precise: profit maximization cannot rationalise a decision that raises growth but wipes out profits forever. Obviously maximizing long-run profits comes under the general idea of maximizing profits.

        “I’d say: ability to control quantity, price and possibly quality while maintaining market share/profit margin/whatever the firm is aiming for.”

        What your talking about there, whether you realize it or not, is predictability. Think about a firm in a very competitive market with no market power in the standard monopoly power sense. It may well be able to control the quantity it sells and hit a market share target. Alternatively a firm in a highly concentrated market might find quantities and hence margins and market share very hard to control. Moreover, the definition of market power you give here provides no explanation for why firms with more ability to “maintain …whatever the firm is aiming for” would choose higher markups.

        “I know you won’t be convinced by anything like this, because no matter what businessmen do they’re really just conforming to the PED analysis.”

        nonsense, For instance if firms in very competitive markets had higher marks ups whilst firms with lots of monopoly power chose lower markups that would certainly not conform to standard analysis. And again. nobody pretends firms directly use undergrad pricing theory to choose prices.

        let’s recap
        Nah, let me do a better recap:

        that’s very disappointing, and rather childish and stupid. Anybody can read the thread and see that you did write that you’d seen no evidence of markups relating to PED in tune with standard theory, I provided some evidence, you said alternative theories could equally fit the data, I asked you to provide one of these alternative theories. At no point did I write anything like “blah blah look you’ve contradicted yourself in some convoluted way”

        “vague data with about a 20% margin of error”

        what sort of a point is that? most people would regard being able to predict actual prices using very simple theory to within a 20% margin of error as a absolute triumph.

        I try to engage with your arguments directly, but arguing with you is like trying to nail jelly to a wall

      • #24 by Unlearningecon on October 15, 2013 - 8:00 pm

        And again. nobody pretends firms directly use undergrad pricing theory to choose prices.

        And again, it ascientific to dismiss what they actually do. Actual scientists only use heuristics like MC = MR when they don’t know more, or when the reality is difficult/impossible to model, or possibly when the heuristic is empirically incredibly accurate (something that is probably not achievable in economics). But we do have an understanding of what goes on inside firms, so why not build on that?

        what sort of a point is that? most people would regard being able to predict actual prices using very simple theory to within a 20% margin of error as a absolute triumph.

        Really? Because given the information he presents: the cost of a razor ($4) and the ‘possible’ maximum price people would pay ($8-12, which he doesn’t explain), it seems you could basically eyeball a price and get an estimate as close as that. The PED processes some relevant information, sure, but it does little more than an educated guess, and looking at institutional factors more comprehensively can help use advance our understanding.

        Btw, I agree that if you had a model that predicted all (or most) variables in the economy with that accuracy, even out of sample, it would be a a good thing. But in this isolated example the ‘prediction’ is just not worth getting excited about.

        that’s very disappointing, and rather childish and stupid.

        Yes, summing up a debate in a way that frames your opponent as being unreasonable is childish. I’m glad you agree.

        Anyone could look at this thread, and my previous posts on the matter, and infer that I am primarily concerned with finding out how economies and therefore firms actually work, and am not a big fan of prediction (economists have proven they are pretty poor at ‘predicting’ anything, despite their focus on this matter). Obviously this depends on how you define ‘prediction’, as I tend to agree with Paul Samuelson, who said that the assumptions and mechanics of a theory cannot truly be said to be different from any ‘predictions’ it makes; in other words, a theory is a collection of testable hypotheses, rather than just a machine that spits out one. (Also see John Bragin’s comment on predictions above).

      • #25 by NicTheNZer on October 18, 2013 - 8:37 am

        “most people would regard being able to predict actual prices using very simple theory to within a 20% margin of error as a absolute triumph.”

        I am pretty sure a very simple theory that ‘prices never change’ will do better than 20% most years, almost everywhere.

  8. #26 by Dinero on October 14, 2013 - 2:58 pm

    In the mass production decreasing demand increases prices as the cost of production per unit increases, possibly leading again to less demand.
    Actually this fits with most people’s experience – go out to buy something popular and many shops have it and the price is low ,but try and find something unusual and you have to pay.

    • #27 by notsneaky on October 19, 2013 - 4:13 pm

      Which is a pretty good way of summarizing why PED matters.

  9. #28 by Luis Enrique on October 16, 2013 - 9:45 am

    what makes you think I am not interested in what firms actually do? It might surprise you to learn I conclude one of my papers with a call for more use of case studies, alongside theory and empirics, in a different context, and that was published in a mainstream economics journal indicating that the supposedly theoclassical editors did not find that call entirely unreasonable either. I try not to argue from authority, but I spent about a decade of my life talking to chief executives and finance directors most every day about what they do, and might just have a firmer grasp of what firms actually do (and of accounting, btw) than you do.

    you keep going on about how firms don’t think about the PED, but the PED is just a theoretical way of presenting the question “what do you think would happen to sales if you changed your price?” and if you don’t think firms care about that question and have some idea of the answer, you are a fool. Firms that are able to raise prices without sales collapsing have monopoly power and a low PED, which is a perfectly useful and sensible analytical framework, which is what I have been trying to get you to see, all along.

    You write that you are interested in understanding the economy, but it looks to me like you are solely interested in rubbishing mainstream economics. If you were interested in understanding the economy, you would show more appreciation of stylised abstract models that help understand the logic of profit maximization under different assumptions about market structures and cost functions etc. and are the building blocks of more complex realistic theories. These models are worth knowing, without denying their shortcomings. You’d also show more appreciation of the drawbacks of case studies.

    “summing up a debate in a way that frames your opponent as being unreasonable is childish.”

    actually I summed up the debate to show you hadn’t get answered my question – it was you, after all, who wrote that other theories could equally well explain the data, and it was hardly unreasonable of me to ask you to present one. Writing “so and so’s theory says firms with more market power will choose higher markups” is not an explanation it is a restatement of the thing to be explained.

    • #29 by Unlearningecon on October 16, 2013 - 11:14 am

      what makes you think I am not interested in what firms actually do? It might surprise you to learn I conclude one of my papers with a call for more use of case studies, alongside theory and empirics, in a different context, and that was published in a mainstream economics journal indicating that the supposedly theoclassical editors did not find that call entirely unreasonable either

      It doesn’t surprise me at all, actually. And I have never used the term “theoclassical”, thanks

      you keep going on about how firms don’t think about the PED, but the PED is just a theoretical way of presenting the question “what do you think would happen to sales if you changed your price?” and if you don’t think firms care about that question and have some idea of the answer, you are a fool. Firms that are able to raise prices without sales collapsing have monopoly power and a low PED, which is a perfectly useful and sensible analytical framework, which is what I have been trying to get you to see, all along.

      Yes, and I’m saying there are more comprehensive ways of answering the question of why firms price how they do. The PED implies that firms would adjust price when demand changes, but they don’t. Why? Economists ad hoc on ‘sticky’ wages/prices to ‘explain’ this, but if you instead use the post-Keynesian framework where a firm is making a long term business plan, set at a particular price based on the consideration of various institutional factors, you would see that everything observed is part of the same process, a process which at least the core neoclassical theories of the firm do not explain.

      It might have been acceptable to explain the PED the way you are at the turn of the 19th century, when economics was just at its birth, or even post-WW2 when it was being fully consolidated. But if economics were advancing, it would have moved past it by now, if only based on the simple empirical observation that firms don’t really vary price with demand.

      You write that you are interested in understanding the economy, but it looks to me like you are solely interested in rubbishing mainstream economics.

      The former implies the latter.

      If you were interested in understanding the economy, you would show more appreciation of stylised abstract models that help understand the logic of profit maximization under different assumptions about market structures and cost functions etc. and are the building blocks of more complex realistic theories.

      Oh well that’s settled then, now that you’ve asserted it.

      How about “I am only interested in models if they help to advance our understanding of the economy, not for their own sake (ie mental masturbation)” ?

  10. #30 by Luis Enrique on October 16, 2013 - 12:12 pm

    “and I’m saying there are more comprehensive ways of answering the question of why firms price how they do.”

    no, you have said more than that, you are claiming econ 101 price theory is useless. I have been arguing it’s useful. If you had just been arguing there’s a need for more comprehensive theories than econ 101, that would have been entirely uncontroversial and would of course been answered by pointing to the array of more comprehensive theories that economists have come up with. If PK theory can add something new and useful, great, but it would only been one of many. How many examples would you like?

    “The PED implies that firms would adjust price when demand changes, but they don’t.”

    what a stupid thing to write. How much evidence of prices responding to demand would you like? I can provide countless papers, but for now ask yourself why Microsoft just cut $100 from Surface prices.

    • #31 by Unlearningecon on October 16, 2013 - 12:40 pm

      no, you have said more than that, you are claiming econ 101 price theory is useless. I have been arguing it’s useful.

      It does not predict anything new or interesting. If you give me the information that econ101 requires (costs, some basic industry characteristics) I can tell you what the price of a good or service will be with the same degree of accuracy as econ101. This is not the hallmark of a good theory.

      If you had just been arguing there’s a need for more comprehensive theories than econ 101, that would have been entirely uncontroversial and would of course been answered by pointing to the array of more comprehensive theories that economists have come up with.

      Many of which use the same concepts I’m arguing against. Why can you not just accept that there might be room for a different perspective to the marginalist one? It’s really quite simple.

      what a stupid thing to write. How much evidence of prices responding to demand would you like? I can provide countless papers, but for now ask yourself why Microsoft just cut $100 from Surface prices.

      Well, yes, if a product absolutely blows then the firm will be forced to cut price, but it’s a rare occurrence, rather than a matter of course as neoclassical theory predicts. I have provided you with statistical evidence that price changes don’t happen much, and that when they do it’s usually as a result of a change in costs. Of course, you haven’t bothered looking at it or responding.

  11. #32 by Luis Enrique on October 16, 2013 - 2:18 pm

    but it does tell us something interesting. it tells us what profit maxization entails when you have given cost and demand curves. Until you learnt th model you did not know that.

    Earlier you were telling me how prediction isn’t so important and referred me to a comment in which somebody said explanation was more basic. The theory provides and explanation

    ” Why can you not just accept that there might be room for a different perspective to the marginalist one? “. I wrote: If PK theory can add something new and useful, great, Being interested in different perspectives however does not entail denying any value to econ 101.

    If you would like to see example of mainstream micro departing from the simple marginalist perspective, have a look at this paper, the papers it cites and then look for papers that cite it. You will see it is a paper about what firms actually do.

    http://qje.oxfordjournals.org/content/110/2/321.full.pdf

    I have a conjecture. When you have been talking about evidence showing prices not responding to demand shocks, you are talking about aggregate demand. That’s why you keep mentioning price stickiness as the mainstream explanation. Of course basic micro 101 with PED and all that is about idiosyncratic demand, and an individual firm’s demand curve, and tells us that if a particular firm’s demand curve shifted (perhaps tastes change or new competitors emerge) they’d change prices.

    I guess you could say that an aggregate demand shock could be seen as a shifting out of each firm’s demand curve, but if you ask me that’s not so obvious because what competing firms do matters to what happens to demand curves, and as you should know if you have done intermediate micro, strategic behaviour between firms is a hard problem.

    I have had another look at table 7 in that PK thesis (is that what you meant?) I cannot see any “demand shock” or anything right hand side variable that proxies for changes in the firm’s demand curve in the regression specification. Perhaps I am missing something. The table seems to show prices responding to a bunch of other things, which of course does not tell us about how prices would respond to a demand shift. .

    If you claim prices do not respond to demand, I can refute that with a sufficient number of examples of prices responding to demand. What happened to the price of Quinoa when it become fashionable, for instance?

    http://www.theguardian.com/commentisfree/2013/jan/16/vegans-stomach-unpalatable-truth-quinoa

    How many papers showing prices responding to demand would you like me to find?

    another example: the price of flights around Wolrd Cup time from UK to Brazil increased sharply last night

    • #33 by Unlearningecon on October 18, 2013 - 2:08 pm

      @NicTheNZer

      Indeed – my time series lecturer pointed out that basic econometric models based on past performance often make predictions close to or even better than the most complex macroeconomic models. As Imre Lakatos pointed out, science is not just about being trivially “correct”, but about discovery.

      @luis

      The theory provides and explanation

      How can the PED explain the process of price formation if the process doesn’t actually, implicitly or explicitly, involve the PED?

      I have had another look at table 7 in that PK thesis (is that what you meant?) I cannot see any “demand shock” or anything right hand side variable that proxies for changes in the firm’s demand curve in the regression specification

      Beta5 is supposed to be how a form responds to a shift in demand (measured as output)

      Let me qualify my general statement about demand and price: I was referring to non-capacity constrained firms, usually manufacturing goods, who try to avoid price changes, particularly decreases.. So no, I am not referring to aggregate demand, at least not with routine variations, as firms generally have spare stock and capacity; even in the long run they are more likely to expand production than put prices up and up (after all, if this weren’t true Apple products would be even more expensive than they already are).

      However, there are important caveats. With agricultural examples like Quinoa, there is indeed an “upward sloping supply curve” due to fixed land; similarly, with positional goods like world cup tickets, where the number is fixed, then yes, price is the only thing that can change. Lastly, even manufacturing firms might be forced to change price if it completely mis-estimate their market, as with Microsoft. However, with established products this is far less likely to happen, and if it did would happen in bursts (such as with new technologies etc.)

      Furthermore, tet’s take stock here. It’s clear you think I’m just being flippant and attacking the poor old mainstream who have been given too much of a badgering recently, and refuse to listen to your counterpoints.

      But this is not so. I quite simply don’t see a use for the neoclassical theories.

      If I were a regulator of, say, energy companies, I would place absolutely zero value in neoclassical theories of the firm. What am I going to use – Cournot, Bertrand? No real world markets behave in that way, and even if they broadly did it would be incredibly difficult or impossible to derive the relevant curves. I’d simply go based on studies of the firms and industry, and would rely on lawyers and statisticians more than economic theory.

      Now, what if I wanted to make an “abstract” model of the economy? Well, first, I tend to think that even the most “abstract” models should be grounded in some sort of observed reality; otherwise, what are they actually modelling? So I would specify it as a model of, say, the UK economy, and would use cost-plus pricing, which is the most commonly observed method, again varying according to industry depending on the exact nature of the model.

      • #34 by NicTheNZer on October 21, 2013 - 2:44 pm

        In my opinion if your not able to do better than trivial models (not to say every model must be better) then you are doing something wrong in the scientific sense however.

        The weather forecasting revolution of numerical computer modelling would not have had much impact if it didn’t perform better than contemporaneous techniques.

  12. #35 by Luis Enrique on October 18, 2013 - 8:13 pm

    Again, PED is “how do you think sales would respond if we changed prices? ” if you think that concept is not involved “explicitly or implicitly” in how prices are chosen you are in cloud cuckoo land. Remember there is a difference between how and whether prices reposed to possibly transitory changes in demand, and how firms chose their price in the first place. As everybody knows there are reasons not to want to muck about with prices, and remember a monopolistic firm would be happy to sell more at its chosen price. Econ 101 is a very simple story that tells us what profit max looks like and impled relationship between markup and PED, then if we observe markups do indeed vary in line with PED then we have an explanation, firms are trying to max profits. It’s all very well saying firms slap a markup on average costs but how is the markup chosen? You mention “various institutional factors” but as yet offer no hint of an explanation as to why these factors, or why market power as you conceive of it, would explain why markups are chosen in the fashion we observe.

    Of course the full richness of reality contains more than a static optimisation problem with two lines on a graph. Yes things like spare capacity, long term thinking etc make a difference. But the basic message of econ 101 that a firm trying to make muchos profits would choose a markup that has a lot to do with how much pricing power it has (its PED) is such an obviously useful part of understanding the economy – even if you then go on to show that firms are often unable to do it with any precision and so do something a bit different, evoutionay learning, whatever – that your denial of any value in it just looks crazy to me.

    Also, ok, sorry I missed what beta5 was. But regressing prices on output is a hopeless way of testing how firms respond to demand shocks. This goes right back to the founding of econometrics and the realisation that you need to identify demand shifters. Otherwise, for example, you could be picking up lower prices causing higher sales. The perils of relying on phd thesis you finding the web.

    And, nicnz, there’s a big difference between predicting/explaining how prices change over time, and what they’re set at in the first place. Getting the levels right to 20% would be far better than I’d dare hope for.

    UE if you were not thinking of AD then I am at a loss why you kept mentioning price stickiness, nobody thinks stickiness stops Rolls Royce choosing a higher markup than Kia.

    Remember prices are nominal. When mainstream macro puzzles over price stickiness it is respect to changing prices so as to neutralise monetary policy. Now reconsider your claim that firms vary output over long run but hold prices. What happens if you hold nominal prices in presence of inflation?

    If you were a regulator and did not think that left to their own devices energy companies would exploit monopoly power, push up margins and make excess profits, why would they need regulating?

    • #36 by Unlearningecon on October 19, 2013 - 4:29 pm

      @luis and notsneaky

      My problems with the PED are these:

      (1) As a general theoretical concept, it has a lot of implications which clearly aren’t true, such as the idea that prices will move often and procyclically (see Blinder on this). It implies that reductions in price will always stimulate increases in demand, even though in Blinder’s study 40% of managers suggested there would be no change. The problem is as I point out in this post: by “estimating” and using the PED, we are assuming the world works a certain way, despite the evidence that contradicts this view. To this end, I’d like to see some studies where the PED was estimated, and then acted on – did what happened after the latter conform to the former?

      (2) It is quite rare for firms to use it with the frequency, and in the way envisaged, by neoclassical economists, with their attempts to model it as a method of cost plus pricing. Notsneaky, I did not pluck this claim out of thin air as you seem to think – I am referring primarily to study’s like Hall and Hitch’s, or more recently Lee’s post-Keynesian Price Theory, who found that the PED was rarely mentioned (and while we’re at it, that the concept of ‘marginal cost’ was alien to many businesspeople). You say that some firms do use PED explicitly, which is a fair cop – you should never say never in social science. However, they use this in far less precise way than neoclassical theories, and Kalecki etc have tried to build theories that take these imperfect guesses into account.

      There is an element of truth to the PED: it obviously process relevant information, and is consistent with some basic observations and common sense. Yet it clearly doesn’t explain the whole story, and this is why I prefer the post-Keynesian/institutionalist perspective: it makes “stickiness”, firm structure and aggregate demand all part of the same story. Firms make a business plan based on a price which is set depending on costs and numerous considerations, which will vary depending on the market and firm. Prices are not market clearing, and so leave room for changes in demand to change quantity, not price – this avoids costly, destructive and unpopular price changes. Luis, you seem to think I am confused between AD and industry demand. But I am not; I am saying that keeping prices relatively constant but varying quantity is the preferred demand-management strategy of many firms, and this is perfectly compatible with a macroeconomic AD story, too.

      Now, if you think these theories should be given a place in economics, that’s fine – we agree.

      Some of the less important points:

      Notsneaky, you question whom I would employ, a mainstream economist or a heterodox nutter (and btw, you deliberately set up the way you asked the question to make the heterodox economist look worse; I could do the opposite, but I won’t). I can tell you that insofar as I would choose the mainstream one, it would be due to them having a degree, you having made it sound like the heterodox one doesn’t. More importantly, I would not want either of them based on their economic-school leanings. I would simply prefer somebody who had knowledge of the industry’s “institutional characteristics”, whether they were a lawyer, manager or even an economist.

      Also, you were referring to a tweet by me when I quoted my lecturer, who was referring to my own business school, and obviously he was just being flippant for the purposes of a joke. It is true that economics has found its way into some managerial schools, particularly in the USA where it is more dominant than here. Clearly economics interacts with the object it is studying, so this is not surprising. However, it is also true that employers have complained about economics graduates and their “narrow” set of skills, so the question of whether this stuff is worthwhile is still open.

      Luis, your claim about regulation of energy companies is perplexing. I can only guess you think that we need neoclassical economics to know that these firms will do the things you mention? Come on.

      • #37 by notsneaky on October 20, 2013 - 12:00 am

        Just the first sentence for now:

        As a general theoretical concept, it has a lot of implications which clearly aren’t true, such as the idea that prices will move often and procyclically (see Blinder on this).

        Neither is a necessary implication of of elasticity pricing. Pro- or counter- cyclicality of prices is an empirical question and PED pricing is agnostic on the subject. It’s simply a matter of what kind of shocks the firms are subject to. If cost shocks dominate prices are counter cyclical. If demand shocks, they might be procyclical (weakly; if demand is isoelastic and marginal cost is locally constant then no change). Note that mark-up pricing works essentially the same way.

        As to the frequency of adjustments, you’re conflating how firms *set* prices with how they *adjust* them. Similar but not the same. A firm may wish not to change its price frequently for whatever reason (menu costs, near rationality, implicit or explicit contracts etc) but you still want to know how it chooses the price it does set and keep. Firms could set prices based on “average” level of demand and marginal cost with the mark up determined by “average” elasticity of demand.

        Which Blinder paper are you referring to specifically?

  13. #38 by Luis Enrique on October 19, 2013 - 8:40 pm

    I’m sorry but you can estimate “how much demand changes when prices do” without enforcing any theory onto the world, it’s a simple question and it’s answer may be there is no such relationship.

    W.r.t implications that aren’t true: only if you place far more weight on the theory than it can bear. My advice to you is: expect less, get more. Two lines on a graph ain’t going to give you a good answer to the question how do firms respond to transitory demand fluctuations. If you try to use your simple static model, and say well if these lines shift like so then prices will change like so, you will, I agree, get an answer that does not happen. You do not need to reject the simple because life is more complex,you can take something from the simple and think further about everything it misss out. This isn’t an original point.

Follow

Get every new post delivered to your Inbox.

Join 1,021 other followers