Heyne, Paul. Microeconomics. 3d ed. Macmillan College Publishing Company: New York, 1994, pp. 119-136.


Economic analysis is basically marginal analysis. Many economists even use the word marginalism to refer to what we have called "the economic way of thinking." Marginal means additional. Economic theory is marginal analysis because it assumes that decisions are always reached by weighing additional costs against additional benefits. Nothing matters in decisionmaking except marginal costs and marginal benefits.


Suppose that the love of your life phones you at 9 P.M., while you're studying desperately for your physics exam the next day. He wants to come over for a couple of hours. You tell him you have to study. He pleads. You say no. He asks plaintively, "Is physics more important than me?" And if you've grasped the economic way of thinking, you respond without hesitation: "Only at the margin."

If that doesn't stop his whining, tell him to enroll next term in an economics class, and go back to your studies. The issue of his value versus the value of physics just doesn't arise in this situation. The question, rather, is whether an additional two hours with him at this time is worth more than an additional two hours with your physics text.

Since you don't want to alienate him, you might explain a little further. Ask him which is more important to him, water or toothpaste. If he answers, "Water," you've got him. For he undoubtedly uses water as if it had no value at all. And he would probably refuse indignantly to pay as little as 5 cents for a drink of water if someone tried to charge him that price in a restaurant. Why? Because it is the value and importance of additional water in his actual situation that affects his behavior, not the value of water in the abstract. What he and you are each willing to give up to obtain a drink of water is not what you would be willing to pay if you were dying of thirst in a desert; it is what you are willing to pay when you're surrounded by faucets and fountains that offer virtually unlimited quantities of water at the turn of your wrist.

He is thinking in terms of "all or nothing." But that just isn't the choice when he phones on the evening before your exam. In fact, that is rarely the choice we face when we're called upon to make decisions. It's usually more of this and less of that versus more of that and less of this. The economic way of thinking rejects the all-or-nothing approach in favor of attention to marginal costs and marginal benefits.


When you pass through the cafeteria line, pick up the tuna lasagna, and pay the cashier $1.90, you incur a cost: the value of whatever opportunity you will have to forgo because you've spent $1.90. Then you take your first bite and suddenly wish you hadn't selected this item. What will be the cost to you of leaving the lasagna on your plate?

It will not be $1.90 or $1.80, if we assume your first bite consumed about ten cents' worth of tuna lasagna. That cost is history. The cost of leaving the lasagna on your plate will be the value of whatever opportunity you forgo by doing so. Do you have a dog that would enjoy pasta with tuna, cheese, and spicy tomato sauce? If so, the cost of leaving the lasagna is the opportunity you forgo (by not asking for a doggie bag) to see your dog's eyes light up and its tail wag.

The price you paid is what economists dismiss as a sunk cost. Sunk costs are irrelevant to economic decisions. Bygones are bygones.

Of course, we must be certain that a cost is really sunk, or fully sunk, before we decide to regard it as irrelevant to decisionmaking. If you were to purchase a new motorcycle and immediately afterward regret your decision, what would be the cost to you of continuing to own the motorcycle? Clearly, you would not be forced to say, "I did it and now I'm stuck." You could resell the motorcycle. By not doing so you would incur a cost (a benefit forgone) equal to its resale value. The genuine sunk cost would therefore be only the difference between what you paid for it and what you can get by selling it. That is the irrelevant part of your cost. In the economist's way of thinking it is no cost at all, for it represents no opportunity for choice. It may be cause for bitter regret and the occasion of some education in the dangers of impulse buying, but it is no longer a cost in any sense relevant to the economics of present decisions.

Yet we all know that people do not consistently reason things out in this way. Many people who made such a purchase and then regretted it would be tempted to retain possession of the motorcycle rather than sell it for substantially less than the original price. They might justify this action by saying "I can't afford to take the loss." But they already took the loss! They made a mistake, and their full loss occurred when they made it. If they nonetheless choose to keep the motorcycle, they are probably practicing self-deception. They persuade themselves that a motorcycle gathering cobwebs in the garage has the same value as the money they paid to put it there, and more value than the opportunities forgone by keeping it there. But the only relevant cost now is the opportunity forgone by not selling.


Let's practice this approach. Suppose you own a television retail store, and one of your suppliers is sponsoring a gigantic Dealers Contest. For every television set you buy (no returns allowed), you receive one day in Las Vegas with all expenses paid. You gleefully order 28 sets, and your wife starts planning your two-week holiday together.

Upon your return from Las Vegas, you begin wondering how you will sell all those television sets. One month later you're still wondering. It seems that none of your customers is interested in that brand or model. You are about ready to give up and store the whole lot in the back workroom.

Then you get an offer from an orphanage in some distant city to take all those sets off your hands for $4000. You know that a businessman can't make money by selling below cost, so you sit down to figure out the cost of the sets. You paid $140 apiece to the supplier. Moreover, you have had them in your store for a month tying up valuable floor space. You borrowed the money to buy them at 12 percent annual interest. You also had various handling costs, which you estimate at $400. And you spent $160 on advertising in a vain effort to move the sets. By estimating $280 as the cost of display space tied up for a month, you arrive at a figure of $4800. You write back to the orphanage that you would be willing to sell the lot at cost for $4800, forgoing any profit on the transaction in the interest of charity. The orphanage replies that $4000 is their top price, since they can get the sets they want somewhere else for that price. But you are a good businessman, you know that losses don't make profits, and you refuse.

You were actually a rather poor businessman. Every one of the "costs" you enumerated in arriving at your total of $4800 was a past expenditure and hence no cost at all. The proper stance for making cost calculations is not looking back to the past, but forward to the future. Your costs, if you sell, will be the opportunities thereby forgone, or what you can get for the sets if you do not sell to the orphanage. You know the market fairly well and you estimate you could get $1120 by selling them for junk. The marginal cost of selling to the orphanage is therefore $1120. Your gain from selling to the orphanage is consequently $2880. Any loss that you're worrying ~,t about should be assigned to experience and the glorious memories of Las Vegas. It is irrelevant for decisionmaking purposes.


The word marginal means in economics exactly what it means in everyday speech: situated on the border or edge. The concept is of fundamental importance in economic thinking, because economic decisions, like all effective decisions, always involve marginal comparisons. That is, they always have to do with movements at the border, with positive or negative additions. What will be the additional, or marginal, cost that results from this decision? And how does it compare with the marginal cost of alternative decisions? If you think about it for a moment, you will discover that opportunity costs are always expected marginal costs. The term marginal cost does no more than bring into strong relief an aspect of opportunity-cost thinking.

 It's important not to get the marginal concept mixed up with the notion of average. You may have no intention of confusing marginal with average; if so, what follows may only plant in your head the seeds of a bad idea. Let's hope it doesn't. A simple production schedule of a hypothetical zerc manufacturer will illustrate the distinction.

Number of
Zercs Produced
Total Cost of Producing Zercs
42 $4200
43 4257
44 4312
45 4365

A little long division reveals that 42 zercs can be produced at an average cost (total cost per unit) of $I00; the average cost is $99 for 43 zercs, $98 for 44, and $97 for 45. A little subtraction reveals, however, that the cost of producing the 43rd zerc is not $99 but $57. The incremental expenditure, or the extra cost, incurred by producing the 43rd zerc, is its marginal cost. The marginal costs of the 44th and 45th zercs are $55 and $53, respectively. It is clear that marginal cost can be more or less than average cost and can even differ substantially from average cost. It should also be clear that for a zerc manufacturer trying to make production decisions, it is the marginal costs that should guide him. Shall we produce more? Or less? Marginal cost is the consequence of action; it should therefore be the guide to action.

Are business people then not interested in average costs? Unless they receive sufficient revenue to cover all their costs they will sustain a loss. They won't willingly commit themselves to any course of action unless they anticipate being able to cover their total costs. They might therefore set up the problem in terms of anticipated production cost per unit against anticipated selling price per unit. But notice that the anticipated costs of any decision are really marginal costs. Marginal cost need not refer to the additional cost of a single unit of output. It could also refer to the additional cost of a batch of output, or the addition to cost expected from a decision regarding an entire process. Decisions are often made in this "lumpy" way.

 For example, no one plans to build a soda-bottling factory expecting to bottle only one case of soda. There are important economies of size in most business operations, so that unless business people see their way clear to producing a large number of units, they won't produce any. They won't enter the business. They won't build the bottling factory at all. The entire decision--build or don't build, build this size plant or that, build in this way or some other way--is a marginal decision at the time it is made. Remember that additions can be very large as well as very small.

Whether or not business people cast their thinking in terms of averages, it is expected marginal costs that guide their decisions. Averages can be looked at after the fact to see how well or poorly things went, and maybe even to learn something about the future if the future can be expected to resemble the past. But this is history again--admittedly an instructive study--whereas economic decisions are always made in the present with an eye to the future.


The Hertz Corporation conducts an annual study of automobile operating costs to determine how much it costs motorists per mile to drive the cars they own. Hertz takes into account depreciation over the life of the car, license fees, insurance premiums, interest on the automobile loan, maintenance, and of course gas and oil. They then announce that it costs so many cents per mile to drive an intermediate-size car if the car is owned four years and driven 10,000 miles per year.

But what does this really tell us? Does it offer any guidance to someone who is trying to decide whether to drive or to take the bus to work each day? Or to someone weighing a commercial airplane against the family car for a vacation trip? Let's suppose you have been asked by your college to drive your car to a conference. You plan to attend the conference whether you drive your own car or not. The college offers you 25 cents a mile. Will it pay you to drive, or should you say no and hitch a ride with someone else? If you go about deciding by trying to calculate whether the cost to you of owning and operating a car is less than 25 cents a mile, you're being foolish. It makes no more sense to speak of the cost per mile of owning a car than to speak of the cost per mile of owning a house. Neither houses nor cars are owned per mile. They're owned per month or per year, but not per mile. If you nonetheless decide to divide the cost of owning by the miles driven, you will come up with numbers that tell you the more you drive, the less it costs. That just isn't so. Each time you drive, you add something to your costs. In order to decide whether or not to drive, you must know the marginal cost of driving.

Costs of purchase, license, insurance, borrowing, plus maintenance and depreciation not due to operation are all unrelated to your decision whether or not to drive. That is why they are irrelevant. The relevant cost is the marginal cost: How much extra will you be out of pocket if you drive? Be sure to include not only the cost of gas but also the costs of oil, tire wear, and mileage-induced repairs. Insofar as cost can be expected to vary proportionately with mileage driven, it can properly be expressed as so many cents per mile. If it is less than 25 cents, as it probably would be, you make money by driving your car. As long as the marginal cost remains less than the price paid by the college, you gain from every additional mile "produced" and "sold."

Are the costs of purchase, license, insurance, borrowing, and time-related depreciation completely irrelevant? Shouldn't you be allowed to cover these costs, too? After all, you have to pay them even if they are not related to the trip you've been asked to undertake.

You are certainly free to ask the college for as high a mileage rate as you choose. And if you have no scruples, you are even free to trot out all your sunk costs and wave them righteously. There is ample precedent for such action. But the one person you don't want to confuse is yourself. Only marginal costs are relevant to your decision, whatever price the college finally agrees to pay you.

Perhaps you've begun to suspect that it's our example that is irrelevant. We're interested, after all, in ordinary business decisions, and whatever may be true of the student driver as entrepreneur, businesses surely have to cover all their costs, not just marginal costs. It would seem so. But it isn't so. There is no more necessity to cover sunk costs in the business world than there is in our case study.

The plain fact is that each year many businesses fail to cover sunk costs. But most of them don't stop operating. We can illustrate by supposing that you bought your car with the intention of driving it for the college. When you made up your mind to become a sort of taxi operator, you hoped to make enough money to pay your way through college. So you purchased a car, the license, and insurance. You probably would not have done so had you expected the college to offer only 25 cents a mile. Maybe you had reason to believe they would pay you 50 cents. Your calculations in these circumstances might have run as follows:

Your problem now is the familiar one of adding apples and oranges. Or even worse, adding apples and velocity. How can you add $8000 to 8 cents per mile and $6 per hour? Obviously you can't.

You could, however, turn all these figures into costs per year, much as Hertz does. The license and insurance are annual costs. The purchase price could be converted into an annual figure by estimating annual depreciation and adding the annual interest charge on your initial outlay of $8000. But you can't state the other costs on an annual basis without knowing how far and how long you'll be driving. You must -(~ anticipate. Remember what we said earlier: the significant costs and benefits in economics are expected costs and benefits. That means they're uncertain. But uncertainty is a fact of life, and if you want to be a student entrepreneur, you'll have to live with it. So you estimate that you'll obtain so many miles and so many hours of business per year. You can then plug in your estimate and obtain numbers for gas and oil, maintenance, and chauffeur services per year.

Now you can do your addition and come up with a figure for the annual cost of doing business. You can then take the mileage estimate you used to calculate gas, oil, and charges for wear and tear, multiply by the rate you expect to be paid, and thereby calculate prospective revenue per year. If the anticipated annual revenue exceeds the anticipated annual cost, you take the plunge and buy a car. If it doesn't, you don't.

So sunk costs are relevant? Of course not. Until you take the plunge, they aren't sunk. They are marginal. They are additions to cost that you are thinking about incurring. That's the essence of being marginal. And as long as they're marginal, they are relevant. But only that long! Until you commit yourself in some way to a business operation, all your costs are marginal. Once you have committed yourself, the situation has obviously changed. If you want to maximize your profits (or minimize your losses, which comes to the same thing), you must produce and sell all those units of output whose anticipated marginal cost is less than the anticipated price to be set by the college, which is the marginal benefit.


But if a business firm doesn't cover its sunk costs, who does? Who pays the bill for mistaken decisions in the past? We're going to look more closely at that issue in Chapter 1 I, when we introduce the concept of the entrepreneur. For now, though, we can say that the costs are paid by the investors who provided the means by which the mistakes were made. They had hoped to profit; but matters failed to turn out as expected, and so they take a loss.

Consider a simple case. Amber Crombie and her husband Fitch pay $250,000 to buy a lot and build a home overlooking beautiful Lake Lilypad. They expect to receive more than $250,000 worth of benefits from living in their comfortable new home with its breathtaking view. Soon after moving in, however, they notice that Lake Lilypad supplies an odor as well as a view. The smell gets progressively worse. Finally, Amber and Fitch decide they can't stand it any longer and will have to move. When they put their house on the market, however, the best offer they get is $60,000, from a couple with chronic nasal congestion. What will happen next?

Amber and Fitch must compare marginal benefits and marginal costs. They will weigh the odor-reduced benefits from continuing to live in their house against the cost of doing so, which is $60,000 forgone. In other words, would they choose to buy this house, knowing what they now know, if the total price were $60,000? If not, they should sell.

What about the other $190,000 they've invested? That's the loss they incurred by paying $250,000 for something that has turned out to be worth only $60,000. They can't avoid that loss by staying in the house, at least not if they themselves place a lower value on the house than what the congested couple is willing to pay. They have sustained a $190,000 reduction in their wealth. And their loss will be even greater, under our assumptions, if they try to avoid the loss by not selling.

Suppose, however, that the Crombies had put up only $75,000 of the original cost, and had taken out a bank mortgage for the remaining $175,000. Who bears the loss in this case? The bank will want the mortgage paid off when the sale goes through; but the sale proceeds of $60,000 won't come close to meeting the Crombies' indebtedness. If they have no other assets with which to satisfy the bank, the bank might decide to prevent the sale. But what is there then to stop the Crombies from defaulting on the mortgage and turning the house over to the bank? Whereupon the bank will sell the house to recover its investment, and get... $60,000 from the nonsmelling couple. In that case the Crombies will have lost $75,000, their original investment, and the bank will have lost $115,000, the difference between what it loaned and what it recovered by foreclosing on the mortgage and selling the property.

There are other possibilities, of course. The Crombies may decide that the (expected future!) loss of honor and credit worthiness is a cost too high to accept, in which case they won't default, but will faithfully continue to pay the bank long after they have abandoned the house on Lake Lilypad. There is a good chance that the Crombies or the bank will try to shift the loss to whoever can be found responsible for the odor, or, if no villain can be located, whoever should have warned the Crombies about the fumes from Lake Lilypad.

Whatever ultimately occurs, the cost of buying the lot and building the house were incurred when these steps were taken. All but $60,000 of that is a sunk cost and a loss. Mortgage payments in this case are the marginal cost of retaining possession, or, if the Crombies sell but continue paying, the marginal cost of retaining their credit rating and their honor.


Let's use this way of looking at costs and decisions to see if it can clarify the controversy over rising medical costs and how to contain them. We will look first at physicians' fees.

People often assert that physicians charge high prices in order to recover the cost of their education. Could this be true? The argument implies that dullards who take an extra year to finish medical school will set higher-than-average fees, and geniuses who breeze through in less than the normal time will set lower-than-average fees. We don't see that occurring. The argument also implies that physicians who went through college and medical school on full scholarships will establish lower fee schedules than those who had to pay their own way. That implication is not confirmed by observation either. Suppose a recent medical school graduate learns that she has only two years to live. Could she raise her fees enough to recover the full cost of her education in her two remaining years of practice? The argument makes no sense. To a practicing physician, the cost of a medical education is entirely in the past and consequently irrelevant to current decisions.

Does the cost of acquiring a medical education have no effect, then, on physicians' fees? The answer is that it affects fees indirectly by influencing the supply of physicians' services. The prospect of many arduous years in preparation deters people from premedical programs. The necessity of forgoing income for all those years and of borrowing money to live and to pay tuition decrease the anticipated attractiveness of a physician's life. The expected future cost of going through medical school thus restricts the number that will eventually supply physicians' services. (This restriction is in addition to others, of course: not everyone who wants to graduate from medical school is able to do so.) All of this ultimately affects the prices physicians charge. But to see exactly how it does so, we must push our analysis further.

The fees that physicians charge for seeing patients are indeed related to costs, but the relevant costs are the costs of seeing patients. What is the cost to physician X of seeing patient A? What opportunity does physician X thereby sacrifice? In the common case today it will probably be an opportunity to care for patient B. What determines the value to the physician of that opportunity? It will be the fee that patient B is prepared to pay.

In other words, people bid for the scarce time of physicians on the basis of their demand for physician care. This demand interacts with the supply of physicians' services to determine the cost of a visit to the doctor, or the price that physicians will be able to charge. The more that physician X can get from seeing other patients, the higher is the cost of tending patient A. Note that this is not determined by the value of the time the physician spent in school, but by the value of the physicians' time to patients. The point is a most important one. If the demand for physicians' services increases but there is no increase in the amount of such services supplied, demanders will simply raise the cost of medical care, essentially by trying to bid it away from other people. In short, we who call for appointments determine the cost of physicians' services.

Physicians are not required to raise their fees when demand increases faster than supply, and many do not. They can instead allow other elements of the cost to rise. "I can't get you in before Thursday." "You can come into the office and wait, if you want to, but the doctor may not be able to see you." The less that fees rise in response to increased demand, the more queuing costs tend to rise. As a result, those who would have been willing to pay more money will tend to surrender some portion of available medical services to those willing to sit for hours reading old magazines.

The reasons for the large increases in recent years in the demand for physicians' services are many. People are living longer, and older people have more ailments. Our expectations of what physicians are able to do have gone up, and so we consult them more often. And increasingly, the people who visit physicians are not required to pay the marginal cost of that visit. The last reason is an especially important one, because it contains a warning against a popular "solution" to the problem of rising medical costs that is really no solution at all.


Any system of medical insurance that pays for physicians' services lowers the cost to patients of obtaining those services--and thereby increases the amount people will want to purchase. It doesn't much matter for the present argument whether the insurance system is a public one financed wholly out of taxes or a private system under which the beneficiary pays all the premiums. So long as patient A's decision to visit the doctor has no discernible effect on the amount patient A must pay, the relevant price is zero. It's true enough that if beneficiaries go to the doctor more often, premiums (or taxes) will have to be increased. But those additional payments will be made by the policyholders (or taxpayers) whether they decide to see their physicians or, alternatively, to stay at home, get plenty of rest, and drink lots of liquids. The additional payments are sunk costs, and so they cannot affect decisions.

 The marginal monetary cost to the patient of a visit to the doctor is zero when insurance pays the entire fee.

Here is the catch in the insurance approach to the problem of medical costs. If we respond to higher costs by offering larger insurance benefits, we subsidize the use of these scarce services. But that increases the demand, bids up the cost, and necessitates yet another increase in benefits. It also makes people ask why we don't do something to keep medical costs down! We will certainly not keep them down as long as our response to the problem is one that pulls them up by increasing the demand for medical services faster than the supply is increasing.


The question of who pays is also important in the case of another, even more important, contributor to rising medical care costs--hospital services. The average cost of a hospital room in 1985 was more than seven times what it was in 1965. It is almost certain that an increase this large would not have happened if the people using the rooms had been the ones who paid the bills. But under a system in which patients' payments for medical service are exactly the same whether they enter the hospital or receive only outpatient care, patients are much more likely to enter the hospital. The care is usually better in the hospital than outside it, and physicians can more easily monitor patients' progress in the hospital. On top of that, insurance often covers the full cost of hospital care, while paying only a part of outpatient care. That's a direct invitation to people to increase the demand for scarce hospital services, and it quite predictably produces a steadily mounting level of room charges.

Hospital administration also demonstrates that sunk costs do have their uses. Suppose a particular hospital adds a 200-bed wing and purchases a lot of sophisticated new laboratory equipment. Once these decisions have been taken, the costs associated with them are sunk costs. But that doesn't mean they can't be useful to the hospital's administrators. If the government and private insurance companies have agreed to make payments to the hospital based on the hospital's cost of providing service to patients, and if the hospital gets to decide what counts as cost, every dollar of those sunk costs will be used. The sunk costs will be "spread over" each patient, according to whatever formula enables the hospital to recover them as quickly as possible without unduly antagonizing those who must pay.

Cost to whom? Benefit to whom? That's always the best way to pose the question of costs and benefits if you want to find out why some policy is being followed. If the benefits from a hospital's ownership of all the most modern equipment accrue largely to its medical staff, if the cost of not pleasing the medical staff falls primarily on the hospital's administrators, if the cost of acquiring that equipment is borne entirely and uncomplainingly by government, insurance companies, or philanthropists, then hospital administrators will purchase expensive equipment even if it is rarely used--and the cost of hospital-care services will soar. In recent years insurers and government have started to pay hospitals a fixed fee for particular services rather than just reimburse the hospitals for whatever costs they claim to have incurred. Hospital administrators as a consequence now have more reason to control costs and less reason to pretend that sunk costs are costs of providing particular services.


The economic analysis of costs is an especially treacherous enterprise for the unwary, because costs often have an ethical and political as well as an economic dimension. Many people seem to believe that sellers have a right to cover their costs, have no right to any price that is significantly above their costs, and are almost surely pursuing some unfair advantage if they price below cost. This way of thinking, in which cost functions as justification, has even infiltrated our laws. Legislated price controls, for example, usually allow for price increases when costs go up but refuse to permit any price hikes that are not justified by higher costs. And foreign firms selling in the United States can be penalized for "dumping," if a government agency determines that they sold in this country at prices "below cost." In circumstances such as these, when costs become a rationalization rather than a genuine reason for decisions, all statements about costs must be inspected for evidence of special pleading.

Prices ought to be closely related to costs, in popular thought, because costs supposedly represent something real and unavoidable. The most enthusiastic advocates of rent control will agree, at least in principle, that landlords should be allowed to increase their rents when the cost of heating fuel goes up. They will never agree--if they did, they wouldn't advocate rent controls--that landlords should be allowed to raise rents merely because the demand for apartments has increased faster than the supply. That would be "gouging," "profiteering," or "a rip-off," because it is unrelated to cost. But such a rental increase is just as surely related to cost as is an increase in response to higher heating bills. When the demand for rental apartments increases, tenants bid against one another for available space, thereby raising the cost to the landlord of renting to any particular tenant. What another tenant would be willing to pay for the third-floor corner apartment in the Hillcrest Arms creates the landlord's marginal cost of continuing to rent to the present occupant. The case seems to be different with higher heating-fuel prices but really is not. The cost of fuel oil is also determined ultimately by the bids of competing users in relationship to the offers of suppliers. Cost is always the product of demand and supply.

When you grumble to your butcher about the high price of hamburger, the butcher will deny all responsibility. "They keep raising the cost to me," he will say. If you were to investigate further to find out who "they" are and why "they" keep raising the butcher's cost, you would eventually discover that "they" are "we"--we who like hamburger and bid against one another for it. To see exactly how this works, imagine a sudden and unexpected increase in the demand for hamburger, occasioned perhaps by splendid summer weather and a surge of backyard cookouts. The first effect will be a depletion of butchers hamburger inventories. When butchers find their hamburger inventories low and the demand continuing strong, they will increase their orders for beef suitable for grinding. With this happening all over the country, meat packers will in turn find their inventories of beef reduced and will try to buy more cattle. But the increased demand for cattle will encounter a relatively inelastic supply curve, and the price of cattle will rise. The packers who must consequently pay more for cattle will increase the price to the butcher, who can then honestly say "they raised the cost," without for a moment suspecting that "they" are in front of the checkout stand, not back at the wholesale meat market. The cost of hamburger in the supermarket is determined by the interactions of demanders and suppliers.


We can easily imagine circumstances under which such a surge in the demand for hamburger would not increase its cost to the butcher. Suppose that the increased demand for hamburger is accompanied by a decreased demand for round steak, chuck roast, and other beef cuts that can be used to make hamburger. The meat packers and the butchers may be able, under those circumstances, to produce the additional hamburger being demanded without increasing their demand for cattle. In that case, there would be no additional bidding for cattle to raise their price, create higher costs for the butcher, and hence produce a higher price for hamburger at the meat counter.

The passage of time can have a similar effect. Suppose the increased demand for hamburger is part of a general, widespread increase in the demand for beef. The demand for cattle will increase at the livestock market, the price per pound will rise, and consumers will end up paying more for beef.1 The competitive bidding of consumers will have increased the price.

But that will trigger a new kind of competitive bidding, some of which will go on within the mind of single livestock growers. The higher price for cattle will make resources more valuable in cattle production than they were previously and will consequently "bid" some resources away from other uses--growing hogs, raising soybeans, feeding chickens, or working less at the business of feeding cattle. The more effective the higher price is in pulling additional resources into cattle growing, the less will the price of cattle (and hence the price of beef) increase as a result of the original increase in its demand. But the supply adjustments that we're now describing take time. That's why supply curves are typically more elastic in the long run than in the short run.

Let's summarize what we've been trying to say in this section. An increase in the demand for any good (hamburger, medical care, rental apartments) will bid up the cost of acquiring the good (its price) to the extent that it does not cause a larger quantity to be supplied. Or looked at from the other side, an increased demand for any good will not raise its price to the extent that suppliers respond by making larger quantities available. The responses of suppliers will depend on the marginal cost of transferring resources out of their current uses into the production of the good for which the demand has increased.

If resources can be shifted at marginal costs only slightly above those already prevailing, the increased demand will lead to a larger output rather than a higher price. But as the supply of such resources is used up, progressively higher

FIGURE 5C Effect of marginal cost on supply curves
prices will have to be offered to cover the rising marginal cost of shifting less suitable resources. Since resources are more mobile, or less rigidly specialized to one use, in the long run than in the short run, the marginal cost of additional output will usually rise less when adequate time has been allowed for adjustments than when suppliers are restricted entirely to those responses that are immediately available to them.

Summaries tend to be both abstract and obscure. You can make the preceding paragraph more concrete and thus more comprehensible by testing each step of the argument against some real-life situation with which you're familiar. (The case of rental housing in the vicinity of your college might be an excellent example on which to practice.)

1. Don't forget that this higher price will also have the effect of decreasing the quantity demanded. Consumers will end up increasing their monthly beef consumption by less than they had originally intended--before their increased demand raised the price.