19 Jan 2016

Pareto Big Macs

Economics 50 Comments

David R. Henderson has a nice post hitting on the sunk cost fallacy in typical discussions over drug pricing. You can see in the comments though that I think this is a trickier topic than perhaps David realizes. (Or, alternatively, David doesn’t realize how much clearer his thinking is on this than the average person’s.)

Let me elaborate on a puzzle I alluded to in the comments. I’m curious to see what you guys think. (BTW, I don’t mean to exclude both of the women who follow this blog; “you guys” is a generic term.)

There’s a general principle from intro to microeconomics that says in a competitive industry, in equilibrium P=MC. So how would we actually apply that in practice to the fast food industry? At the point at which the burgers are already made and sitting on the back warmer, what’s the marginal cost to the firm of the worker picking up the burger and handing it to a customer? 5 cents?

So, in an efficient fast food industry, burgers should be priced at 5 cents. Don’t you dare say that the firm needs to charge at least enough to cover average costs, because (as David points out) that involves a sunk cost fallacy…

Something is obviously not right in the above. But I’m curious to see how you guys would unpack it. If you want to say, “I don’t trust them there textbooks with their funny graphs!” OK fine, but ideally I’d like you to solve it within the world of standard textbook micro, since presumably that can be done.

50 Responses to “Pareto Big Macs”

  1. Todd Kuipers says:

    Ok – this might be missing the point…

    P=MC means that the cost of the last produced item will be equal to its sale price.

    It’s still worth mentioning AC vs MC – most equipment isn’t sunk cost. It’s likely on lease, financed or has significant resale value or salvage recovery value. At 5 cents it’s way better to load your kitchen on a truck and sell it in another town or state.

    • Bob Murphy says:

      Todd Kuipers wrote:

      At 5 cents it’s way better to load your kitchen on a truck and sell it in another town or state.

      One doesn’t preclude the other. I can tell my employees to hand out the burgers for a nickel apiece, and then 2 minutes later tell them to load the kitchen onto a truck. Selling the burgers at a nickel doesn’t carry an opportunity cost of making me forego the kitchen sale. Remember, I’m saying if we analyze the situation at the moment when there are a bunch of already prepared hamburgers sitting in the warmer, waiting for customers to walk up. What’s the opportunity cost of selling those particular burgers?

      Or if that’s screwing you up, picture a middleman who buys fur coats at $50 each and then sells them from his garage to people who walk up. Once he has 10 coats in his possession, the marginal cost of selling one is what?

      • Matt M says:

        I’m not sure how to explain this in econ terms – but the burgers are only pre-made because there is quite a lot of certainty that someone will demand them (keep in mind the original fast food model was to make food to order, until they realized they could save time by pre-making the most popular items).

        If you think fast food is trippy though – what about digital goods? The marginal cost of allowing one extra person to download your book/movie/game is zero.

  2. Kevin L says:

    My first thought was to consider the warmer as an intermediate firm in its own right. But I don’t think that helps. Maybe MC should include opportunity costs. If I sell the burger for 5 cents when minutes later it would sell for 50 cents, that’s a real cost, if not a cost of production. That also explains why Panera gives away baked items before closing: there’s no opportunity forgone, so P does equal MC in the textbook sense.

  3. guest says:

    “… what’s the marginal cost to the firm of the worker picking up the burger and handing it to a customer? 5 cents?”

    The marginal unit isn’t “the worker picking up the burger and handing it to a customer”, but, rather, the cost per hour to hire him.

    So, the marginal cost of the worker handing a burger to a customer is actually zero.

    Toward a Reconstruction of Utility and Welfare Economics
    II – Utility Theory
    Ordinal Marginal Utility and “Total Utility”
    https://mises.org/library/toward-reconstruction-utility-and-welfare-economics-0#2a

    “In human action, “marginal” refers not to an infinitely small unit, but to the relevant unit. Any unit relevant to a particular action is marginal. For example, if we are dealing in a specific situation with single eggs, then each egg is the unit; if we are dealing in terms of six-egg cartons, then each six-egg carton is the unit. In either case, we can speak of a marginal utility. In the former case, we deal with the “marginal utility of an egg” with various supplies of eggs; in the latter, with the “marginal utility of cartons” whatever the supply of cartons of eggs. Both utilities are marginal. In no sense is one utility a “total” of the other.”

    *Takes a bow*

  4. Tel says:

    Shall I give the David Henderson approach a go with the bank and my mortgage? I mean, it really doesn’t matter if I pay the loan back or anything, because hey bank, that’s a sunk cost… don’t even think about accounting for this on the basis of making a return or anything.

    Chance of success on that line? Hmmm…

  5. Tel says:

    … and they conceal evidence suggesting their drugs may be less effective than advertised, or possibly even harmful [DRH: hard to believe].

    I think David is a bit slack on his research there. Not only is it believable from an economic viewpoint (yes incentives really do matter) but it’s been soundly demonstrated by statistical analysis that publishers have a negative attitude to negative results.

    So much so, that there’s a big push to register all clinical trials to ensure they cannot just go missing after the fact. The registration process is deliberately structured as to be difficult to game.

    https://en.wikipedia.org/wiki/Clinical_trials_registry

    Put it this way, suppose you had a bet with Krugman, where he is allowed to just forget about the bet at any future time should he find that convenient… seems like an attractive deal, right?

  6. Strat says:

    I’ll take a stab at it:

    For the Big Mac seller,
    Price = Opportunity Cost.

    If Opportunity Costs are greater than Marginal Costs, the Big Mac Seller(s) will Produce more.

    At some Point Opportunity Costs will Decrease, and Marginal Costs Increase.

    Thus in an efficient market,
    Price = Opportunity Costs = Marginal costs

    At any individual point in time, Price and Marginal Costs could be different (e.g. Digital Goods, Unique Items)

    But even in efficient Digital goods markets, we would expect that Price (returns from advertising revenue) equal the costs of production (time & effort), adjusted for risk, and interest.

    Otherwise more firms would enter the market, splitting audiences (and thus advertising revenue), while increasing the prices of factors of production (Top of the Funnel Marketing, Wages for Content Producers etc.)

  7. Maurizio says:

    P= MC is a relation that holds at the moment they decide how much to produce. They produce the quantity such that P = MC.

    Why are you assuming it holds when they have already produced?

    • marris says:

      Exactly right. The marginal cost of 5 cents is calculated over only *part of* the production period. When the producer decides to produce a marginal burger, he calculates his (forward looking) expected marginal cost. If you told him that his marginal revenue will be 5 cents, then he would decide to not produce any burgers. This producer would leave the market and the market would not be in equilibrium.

    • Craw says:

      Exactly.

      Imagine suddenly a new law mandating $50,00 an hour that magically comes into effect after the big mac is cooked but before the server hands it over and collects the price. Suddenly it’s cheaper to not sell the big mac, so it won’t be sold.

    • Scott says:

      No, I do not think this is right.

      The marginal cost is the cost to produce just the last ‘increment’ of burger. The whole point is to see sunk costs for what they are — sunk, not something you get to ‘spread out’ over the units as you produce them, just because that suits your cost accounting as a producer.

      The Austrian criticism is valid, I think (that P=MC is invalid, for a number of reasons…) Which is what I think Bob is trying to point out. But there is a basic insight to the concept, which I think this way of thinking messes up. Guest’s argument is right, I think, just for basically trying to see it through the lens the way it was meant. What Bob is doing is a sort of reductio-ad-absurdum.

      For example, why isn’t the MC equal to the energy required to volatilize the last molecule of burger grease off the employee’s hand (maybe 1 X 10ˆ-12 joules, which at current energy prices is $1X10ˆ-15, or some other absurdity…)

      • Grane Peer says:

        Hey P=MC is just as valid as calling zero “normal profit”

        I think the point of this was about looking at future costs instead of past costs. By looking at the finished burgers as sunk you are forced to reason your prices by the profit necessary to maintain future production. I don’t know what the MC would be but I think it has to reflect the costs of producing additional run of burgers. I’m guessing a large number of firms in equilibrium are meant to run for more than a day.

  8. Tel says:

    Something is obviously not right in the above. But I’m curious to see how you guys would unpack it. If you want to say, “I don’t trust them there textbooks with their funny graphs!” OK fine, but ideally I’d like you to solve it within the world of standard textbook micro, since presumably that can be done.

    I’d certainly be very careful trusting any economist to handle business accounting, and after reading the above I’d say probably best to quietly cross David Henderson off that already short list. What I’ve noticed is that so few economists are even familiar with common accounting practices. They almost look at it as a the dopey cousin who we keep around to be nice and all but no one listens to what he has to say.

    The point of keeping accounts is to discover important questions like profitability and the most important part of this is being able to group expenses transactions along with sales transactions to get a nett profit or loss. The more fine-grained you can do this, the better you can understand your business activities.

    There are various ways to handle this, but you can see that with any manufacturing or production industry, you need to track all of the cost components, including parts and labor, and then you need to build in amortized fixed costs like rent. If a given product line is not profitable for your business, then you drop the product line (or change strategy, higher prices, more advertising, whatever). Now in the immediate case you still might be pushed into a position where selling at a loss is better than throwing the stuff away… but sure as heck after taking a few big losses the business isn’t going to make any more of those.

    That’s the whole idea of accounting… to tell you what you are doing right or wrong.

    I’ll also point out that the basic economic presumption is that a competitive market is the thing that prevents producers from price gouging, and you don’t get competition in an industry by driving every producer into bankruptcy. This is similar to those price controlled apartments. For an existing stock of apartments, government imposed cheap rent price controls will (temporarily) make housing more affordable (owners are stuck with a sunk cost so they have to take it on the chin), but they also drive away the investment money and discourage anyone from producing more housing stock. In the medium to longer term supply dries up and you get shortages.

    This is also known as “The Hold-Up Problem” and it doesn’t seem like the sort of thing a respectable free market economist would advocate.

  9. Toby says:

    Hi Bob,

    The problem doesn’t seem to be very well posed and I don’t think it follows that in an efficient fast food industry we’d observe burgers to be priced at that marginal cost of 5 cents. I am pretty sure that we would not be observing any fast food industry if that were the case.

    I recall that the textbook account is that a firm operating in a perfectly competitive market has to at least expect to be able to cover its costs or it will simply not enter or close up shop and cease production. At this point, though, the (fixed) cost, that you and David seem to have in mind, are not sunk yet.

    I believe, however, that the practice in the fast food industry is to pay men and women by the hour and hire them for fixed hours. The cost of labor is, therefore, also already sunk (provided that rights are perfectly enforced and without cost) at that point. The marginal cost, at that point, of picking up the burgers and handing it to the customer is at least zero if not negative.

    I would argue that it is negative, because the customer will already have given something in exchange at that point for the burger. This seems to be standard practice in this industry: first the customer pays and then the customer receives his or her order. Not handing the burger to the customer means having to return what was given in exchange (again provided that rights are perfectly and costlessly enforced). It is negative, of course, under the assumption that this is something that you’d wish to keep more than you’d wish to keep the burger.

    This, though, is not the relevant for marginal cost for what price to charge. The conclusion that this is the price an efficient fast food industry, therefore, would charge does not follow. It really depends on the timing of the model that best represents the fast food industry. Or to stay within undergrad microeconomics: it depends on the decisions made in the short run in the fast food industry. The example here seems to be the very very very short run.

    To conclude, I think that the problem is ill-posed as it needs more assumptions. The marginal cost is most likely negative at that point in time. This, however, is not the marginal cost relevant for the pricing decision. If it were, then there would be no fast food industry to speak of.

    Cheers,

    Toby

  10. skylien says:

    Well, obviously costs have nothing to do with what you are able to ask from customers to pay for your product. Thinking otherwise would be believing in some form of cost theory of value. Of course it is the other way around. It is the total expected sales revenue from selling a product in the future that determines how much costs you are willing to bear now for creating the product.

    So the burgers wouldn’t exist if they originally exected to sell them for only 5cents. So I also think it is not wrong to say later on, we want to cover our R&D costs by charging 1000 per pill, because that was the plan from the beginning. The question is can they? If they can they should.

    • skylien says:

      *exected* should be *expected*

  11. Harold says:

    Oh how difficult it is to describe a difference that appears to be obvious, but you cannot pin down. Sunk cost is a fallacy, yet clearly we must consider the cost to produce new goods.

    “At the point at which the burgers are already made and sitting on the back warmer, what’s the marginal cost to the firm ”

    As Kevin L points out, at the end of the day bakers may discount heavily to get some return on the sunk costs of the already baked goods. These goods will spoil and become worthless and are considered a sunk cost, whereas goods already baked earlier in the day are not – they are not sold at knock down prices. So if we can work out why the last buns baked are treated differently to the first buns, we should have the answer.

    At the end of the day, there are no plans to bake new buns. If we sell buns earlier in the day for 5c these will also be the last to be sold, since there will be no incentive to replace them with new ones. If there is no plan to replace the buns, we might as well sell them off cheap. If the pharmaceutical company had no plans to develop any new drugs, they should get what they could for the existing drug. This makes sound economic sense.

    In a competitive market, every baker has the same costs. If one baker could produce cheaper, they would cut prices and take all the business. However, there would be nothing to stop all the other bakers taking the same cost cutting measure, but each would have to make the same investment as the first.

    For drugs this is not the case. One company makes all the investment in research, then any other company can make the drug without having made that investment. It is like the first baker investing in a new oven, then all the other bakers using it. This puts the first company at a huge disadvantage. There must be some mechanism to allow the first company to avoid that disadvantage if we want the new oven. For ovens it is simple, not so for drugs.

    Several routes are possible – we could lower costs to below what the other companies need to spend to copy the drug, or we must raise profits so the costs are recovered, or we somehow charge other companies to make the drug (as the baker could charge ather bakers to use the oven).

    Reducing costs. If we did away with licencing and compulsory testing, it is just about possible that drugs could be made cheap enough to make a profit before other companies stepped in. This is likely to lead to lots of other problems – not least uncertainty over which drug is a good drug.
    Raising profits. The current fix is to impose high costs in licencing and regulation, which means we have good drugs. We then guarantee a monopoly for a fixed period, over which time we know the good drug will be over-priced and under-produced.
    Charge to use. This is like one company licencing the IP to another company. This can happen at present, but does not usually for new drugs. More can apparently be made from the monopoly than from the licenses. I am not sure why this is so. This would seem to be the best solution, but I do not know how we could make it come about. It is like there being one baker with a huge oven, he only uses a small part of it, but he does not rent out the rest of his oven space to other bakers to make money, instead he charges a fortune for his own cakes.

    • guest says:

      “It is like the first baker investing in a new oven, then all the other bakers using it. This puts the first company at a huge disadvantage. There must be some mechanism to allow the first company to avoid that disadvantage if we want the new oven.”

      I submit, therefore, that we don’t really want the new oven unless it’s at someone else’s expense. In which case, the new oven ought *not* be invented.

      When a baker believes he can hide his new oven long enough to make a sufficient amount of profit, or more, he will buy it.

      Same with drugs, access to which is also not a right.

      • Harold says:

        I don’t see who’s expense this is at. The oven owner gets his fee, the other bakers choose to pay to use the oven and the customers get buns to buy. Where is the expense?

        • guest says:

          If the oven owner gets his fee, how does that put “the first company at a huge disadvantage”, as per your scenario?

          Crisis averted thanks to the profit motive.

          • Harold says:

            I see -I was not clear. I meant if all the other bakers used his oven for free he would be at a disadvantage. If he can charge, then everyone is a winner. Since the oven is in his shop he can indeed easily restrict access and recieve his fee.

            Comparing this to drugs or IP in general, there is no way to prevent other companies using your new drug discovery for free. Secrecy used to be an option, but modern analysis can identify the molecule and reproduce it.

            So we need some way to make IP the same as the oven – to stop anyone else using it without the owners permission. Hence patents.

            • guest says:

              “Comparing this to drugs or IP in general, there is no way to prevent other companies using your new drug discovery for free.”

              OK, so then I resubmit that we don’t really want the new drugs unless they are at someone else’s expense. In which case the new drugs ought *not* be invented.

              😛

              Eventually, someone will want them bad enough.

              • Harold says:

                No, that is wrong.

                Do you agree that the whoile ciommunity benefits from the oven? This is only because the baker can control access to it. If he could not do so, he will not butyi t and there will be no oven.

                The new drug is the oven. Only if the owner can control access to it will it ever be developed. Why is this at someone elses expense whan the oven was not?

              • guest says:

                “Do you agree that the whoile ciommunity benefits from the oven?”

                No. Specific people benefit from the oven. I reject the concept of externalities because it is a category error to think of economics in terms of aggregates.

                “Why is this at someone elses expense whan the oven was not?”

                In the case of the oven you said: ” I meant if all the other bakers used his oven for free he would be at a disadvantage. If he can charge, then everyone is a winner.”

                In the case of drugs you said: “So we need some way to make IP the same as the oven – to stop anyone else using it without the owners permission. Hence patents.”

                No one is being robbed when other bakers rent the oven from the owner, and as soon as they get enough money they can buy their own.

                Patents forcibly prevent people from using their own property as they see fit.

                If someone owns the ingredients to a drug, then they logically own any conceivable configuration of those particular ingredients, and may therefore sell that drug without permission from anyone else.

              • Harold says:

                You seem to be missing my point. I think we agree that when the Baker buys his oven everybody as an individual gains or remains the same, since he can rent out its use to other bakers.

                “No one is being robbed when other bakers rent the oven from the owner, and as soon as they get enough money they can buy their own.”

                The baker forcibly prevents others from coming onto his land and using his oven for free. Therefore he can charge for its use. If it was out in public he could not prevent others from using it, therefore he would not buy it in the first place.

                Similarly for a new drug. Just as the baker will never buy the oven unless he can restrict access to it (forcibly), the drug company will not develop a new drug unless they can restrict access to the knowledge required to make and sell it.

                “Patents forcibly prevent people from using their own property as they see fit.”

                Well, it is not so clear cut. The information to make the new drug is like the oven. Without patents it is like the baker having his oven out in the open – he cannot prevent others using it for free. Therefore the new drug will never be developed. The question is whether the information that he first company has gained by developing and testuing the drug is their property.

                It is not so simple, because the patent ought not stop anyone else who developed the new drug independently. Thus the patent might prevent someone from using their own property. If we

              • guest says:

                “the drug company will not develop a new drug unless they can restrict access to the knowledge required to make and sell it.”

                Or unless someone wants the drug bad enough to pay the high price.

                Further, as you noted, analysis equipment is plentiful, so there’s competition that can drive the price down.

                And then there’s the regulations that keep drugs off the market.

                Drug companies not being able to keep their ingredients secret is not a problem.

                “The information to make the new drug is like the oven.”

                Information isn’t what a lot of people think it is.

                If I can’t read something because it’s in another language, it’s not information to me.

                This scenario exemplifies the fact that information is really just meaning that the observer *himself* pours into what is seen or heard.

                In other words, information is not created by the communicator, but by the receiver.

                (The communicator is “receiving” the communication as he’s recording/conveying it.)

                So, information can never logically be transferred, and therefore is not something that can be bought or sold.

                What we are buying when we “buy” information is actually the service of recording or conveying symbols or sounds.

                They remain mere symbols and sounds until the receiver pours meaning into them, at which point they become information.

              • Anonymous says:

                “Similarly for a new drug. Just as the baker will never buy the oven unless he can restrict access to it (forcibly), the drug company will not develop a new drug unless they can restrict access to the knowledge required to make and sell it.”

                That is clearly wrong considering drugs were produced before patents. You can argue that patents bring more drugs to market (I’d disagree and ask you to show me some studies that demonstrate that), but you’re obviously wrong to say drugs would not be produced without patents.

              • Harold says:

                Anonymous – See my point about analysis – secrecy is no longer an option. If I develped a drug in 1880 it was impossible for others to find out what it was. Now it is possible and relatively easily.

                Guest:
                Harold:“the drug company will not develop a new drug unless they can restrict access to the knowledge required to make and sell it.”

                Guest: “Or unless someone wants the drug bad enough to pay the high price.”

                There will not be a high price (or only for a very short time). As soon as competitors see the high price they will copy the drug cheaply.

                Say company A spends $10B on research to identify a molecule that cures cancer. This drug could easily make 100 times the development costs, and lots of cancer sufferers will benefit.

                Once this molecule is sold, another company can analyse it and identify the molecule for $1M. The second company is now in just as good a position to sell the drug as company A that discovered it. But they pent only 0.01% what company A spent.

                Great – because competition will drive the price down to a little over production costs. All those cancer sufferers will benefit for less.

                Not so great, because at that price the first company will never make $10b. Why would the first company bother to spend £10B in the first place? There will not be a molecule for the second company to copy.

                All those cancer sufferers die of cancer after all.

                Can you explain how company A can make a profit unless it can restrict others from using the knowledge of the identity of the molecule?

  12. Don Boudreaux says:

    This issue is indeed one that is set-up by standard economic analysis to be confusing. The reason is that distinguishing between “short-run” and “long-run” is a far too inadequate way of incorporating considerations of time into the economic analysis of production. Much of this confusion would disappear were more economists familiar with Armen Alchian’s brilliant, but relatively unknown, 1959 paper “Costs and Outputs.” (Here’s a link to a draft version: http://www.rand.org/content/dam/rand/pubs/papers/2008/P1449.pdf )

    Alchian, apparently, had gotten this article accepted for publication in the American Economic Review but then pulled it (!) so that he could publish it in a festschrift for Bernard Haley (who, I believe, was Alchian’s professor at Stanford). It’s a shame that this paper never appeared in the AER, for it likely would have had far greater influence.

  13. Dan W. says:

    As a teenager I learned that if one hangs out at the fast food joint until closing one can go home with a variety of “free” food. Of course the marginal utility of the 3rd cold hamburger at 1 am is about what you would think it would be.

    • ejk says:

      You have also have to be there until 1 am. The opportunity to do something more productive or pleasant is lost. Of course, you’ve already decided the path of greatest value to you is that 3rd cold hamburger.

  14. Innocent says:

    If you are going to say that ‘sunk costs’ don’t matter then movies should be free right? They have been made and have no value once they are on film. Especially since now a days digital only costs a couple cents of electricity to move and down load. Right? Why should the internet cost more than a couple cents to access, after all there was a sunk cost in connecting things but that is over and done with.

  15. RPLong says:

    “At the point at which the burgers are already made and sitting on the back warmer, what’s the marginal cost to the firm of the worker picking up the burger and handing it to a customer? 5 cents?”

    For various reasons, this number is too low. Having spent a fair amount of time working in fast food, I can tell you how serious management can be about preventing idle burgers from stacking up. This happens in anticipation of peak traffic periods, sure, but in general a restaurant does not keep a stock of food on the warmers. Reasons for this include:

    1 – Asset depreciation; a fresh burger is worth far, far more than a stale leftover. After a set amount of time, established by corporate, an idle burger must be thrown out.

    2 – Shared assets; What happens when you make 3 extra burgers, but then your next five customers order chicken sandwiches, which utilize the same buns and toppings, but a different patty? Product waste, that’s what.

    3 – Opportunity costs; Management would often rather have employees spend their down time doing dishes or slicing tomatoes than cooking up a stock of burgers. There is plenty of other work to be done.

    4 – Unintended consequences, Part One; employees – especially teenage employees – will develop a habit of cooking too much food if they spot an opportunity to eat that which must be thrown away.

    Now, you could say, “Sure, but assuming the burgers have already been cooked, the firm may as well set the price of the burger at 5 cents or something, rather than lose everything on an idle burger. This brings me to…

    5 – Unintended consequences, Part Two; If customers spot an opportunity to get an idle burger at a deep discount, they will develop a preference for stale burgers. They might even try to game the system, sending their friends to order a large number of burgers, cancel their order, and then they all show up several minutes later as “new customers” interested in the old food.

    More generally, since the firm is monopolistic, albeit not monopolist, it will avoid engaging in any business practice that would put downward pressure on its own prices, even when that means pricing above marginal cost. It’s monopoly behavior.

    • Silas Barta says:

      Right, per my other comment, you *can* fit these into a framework whereby P is equal to the appropriate measure of MC … but you’re not really using the core insights of marginal utility analysis anymore, you’re just tabulating “normal” costs and benefits over a bulk of units, where your analysis never actually looks at the marginal unit as the driver of anything.

    • Tel says:

      Yeah, for a given style of burger, within a local area easily reachable at lunch time, that burger joint is a monopoly… and that’s by design.

      However, usually there’s other alternatives: a pie shop, fish and chips, a Chinese takeaway, Indian curry bar, etc. not too far away.

      Then the burger joint wants to get maximum leverage based on the preferences of the people in that area (some of them really lover burgers), and that’s certainly going to be higher than marginal cost, but it would be complex to figure out how much higher. This is all kind of irrelevant because the real issue is how to correctly calculate what marginal cost really is… and the accountants have already spent decades polishing their methodology.

  16. Harold says:

    To condense the main on-topic point from my rather longer post above, if you plan to make more burgers you need to get full value of all the costs from the current batch to fund the next batch. If you are not going to make more burgers, they are a sunk cost and you should get whatever you can for them.

  17. Craw says:

    In Henderson’s example, consider the question “if it cost 5 cents to make and sell a pill, would they still make it if they could only sell it for 10 cents?”. The answer in standard econ, and common sense, is yes. The reason we might support a monopoly price is to provide an incentive for the *next* drug whose development costs are still afloat.

  18. Levi Russell says:

    Even though Don Boudreaux has already chimed in here, I’ll go ahead a post a link to his response to Baker’s piece:

    http://cafehayek.com/2016/01/time-incentives-expectations-and-prices.html

    Because the textbooks have more Stiglitz than Demsetz, we can’t have nice things.

    😛

  19. baconbacon says:

    I skimmed through the answers and everyone seems to be dancing around the standard reasoning.

    Econ 101- Producers want the highest price that they can get for their product, which is based on the consumers willingness to pay. Consumers want to pay the lowest price for the product which is based on producers willingness to produce. Convergence happens near marginal cost because a profit opportunity exists if firms are charging substantially above the MC.

    The MC then becomes a function of how expensive it is to enter, and crucially maintain a presence in, the market.

    The flaw is here

    “There’s a general principle from intro to microeconomics that says in a competitive industry, in equilibrium P=MC”

    This principle is for a PERFECTLY COMPETITIVE INDUSTRY- which means no barriers to entry. If burgers just materialized fully cooked and wrapped then the MC would be something like 5 cents to hire a person to hand them out. Since there are barriers to entry the price of burgers approaches, but does not realize, the marginal cost of handing out each burger.

    Henderson points this out in his last paragraph

    “By the way, if we could completely end government power over the introduction of drugs, maybe a system without patents could work, although I doubt it. But with legislation that makes drugs illegal before the FDA gives permission, and with the requirements for permission imposing costs that average over $2.5 billion per drug brought to market, there’s no way.”

    • marris says:

      The bad assumption is not perfect competition, but equilibrium. If some genie showed up and produced burgers, then the equilibrium would shift since other producers would go out of business and leave the industry. The new equilibrium price would become five cents.

      In a world with no genie, any producer or set of producers’ attempt to charge less than five cents would put them out of business. This would remove them from the market and leave only the producers would “took” the market clearing price.

  20. Capt. J Parker says:

    P=MC will maximize revenue under the condition that the market sets the price. P=MC does not guarantee that total revenue will cover total cost. Total cost includes the cost of capital. If a firm cannot pay a market rate of return on capital it will fail and capital will be redeployed by a firm that can use it more efficiently.

  21. scott says:

    Maybe I’m just repeating guest’s comment above, but clearly the answer has to be that 5 cents is not the marginal cost to produce the burger. The marginal cost would have to at least include the materials, fuel, wages, etc. (but maybe not capital and other overhead, etc.)

    Otherwise, it really would cost 0-5 cents (which I think a lot of other commenters are confirming by talking about hanging out at fast food places late at night & getting free cold burgers.) Once the burger was already cooked, the marginal cost really was approximately zero.

    If the burger place has five burgers sitting on the warmer, it won’t just give one away for 5 cents, because that would reduce its inventory & it would have to replace the burger to return to its existing state. Only when it doesn’t expect to replace them (at closing) does the marginal price go to zero.

    Or something like that. But I’m just guessing….

    • guest says:

      ” Once the burger was already cooked, the marginal cost really was approximately zero.”

      The reason it’s zero is because the company isn’t hiring people to hand cooked burgers to customers.

      It’s not worth it, in the company’s estimation, to hire someone specifically for that purpose, so they won’t pay anything at all for it.

      Rather, they are hiring people to follow the manager’s orders an hour at a time, which may include handing burgers to customers.

  22. Bill says:

    It’s important to distinguish between historical expenses (regardless of what time period they occurred) and opportunity cost. The opportunity cost of handing a hamburger to customer A is reflected in how much the next customer is willing to pay for that hamburger.

    On a related point, gas stations’ pricing strategies are influenced by what customers are willing to pay AND what it costs to replace fuel that is sold. These factors weigh on opportunity cost, NOT the price, i.e., the historical cost, the gas station paid for the fuel that is in the storage tank.

  23. Silas Barta says:

    Yeah, this is exactly why I don’t like repeating P=MC, because it requires so many assumptions that you’re no longer talking about something resembling what we actually deal with in practice.

  24. Maurizio says:

    Note to everybody: P = MC also holds for price makers (monopoly). i.e. a monopoly produces the quantity such that P = MC.

    The reason is that MC contains not only expenses but also forgone revenues. If you produce one unit more, the price decreases, and thus you have a forgone revenue which you need to count in MC.

    For more details see here: http://tinyurl.com/gta6mao

  25. Transformer says:

    At the point of sale almost all the costs of production are fixed (sunk) costs. if the marginal cost of selling an already produced burger is less than 5c then it makes sense to sell it at 5c if that is going rate and it is of less value to keep it.

    Further back in the production cycle of the burger more of the costs of production (labor, depreciation of equipment from use etc) become variable costs, and will be taken onto account when deciding how much to produce and will probably be well above 5c.

    For a big chain like McDonald’s its easy to see how even the decision to open one additional store containing lots of “fixed” capital would be seen as a variable cost that will directly affect the marginal costs of burger production.

    • guest says:

      “At the point of sale almost all the costs of production are fixed (sunk) costs. if the marginal cost of selling an already produced burger is less than 5c then it makes sense to sell it at 5c if that is going rate and it is of less value to keep it.”

      There is no marginal cost of selling an already produced burger because that’s not an economically relevant act (stay with me, here) with respect to the operation of the fast food business.

      There’s no need to break down the costs further than the hourly wage that must be paid. The relevant marginal unit in play here is the hourly wage. At least as far as employees go.

      Nobody is being paid for the specific act of handing out burgers.

      By the way, this is also why it’s a category error for job interviewers to expect interviewees to come up with a number as to how much money they made their previous employer, if they were part of a team / department.

      The team manager is responsible for making the money, and the individual members are delegated tasks that make the department (a relevent unit from the perspective of the individual business owner; individual members of which are expendable) profitable or not.

      This question cannot be answered with a specific number because the individual team member is not a relevant marginal unit to the owner.

      For this type of employee, the interviewer should already have an understanding of what skills will be the most profitable to the department he is hiring for, and should only be concerned about skills.

      (Try not to read “people don’t matter” into that. The purpose of a business is to make the owner a profit, as the owner understands it. Go be someone’s friend if you want to matter to someone.

      (And when you can respect the profit-centric nature of business, maybe you can be someone’s friend *while* trying to make as much profit off of each other as possible – which is entirely possible and rewarding, and a relationship that can open doors for the both of you.)

Leave a Reply