# 17.2: Consumers' and Producers' Surplus

- Last updated

- Save as PDF

- Page ID
- 58529

Society’s resource allocation problem is an especially important optimization problem. It is an easy problem to envision. Pile up of all of society’s factors of production and then ask, "How should we use these resources? What should we make? How much of each product should be produced? How should we distribute the output?" These are questions about resource allocation.

An important idea is that of a constraint. Needs and wants by consumers far outstrip available resources. More of one means less of other goods and services.

The previous section showed how supply and demand establishes an equilibrium price and output. The latter is the market system’s answer to the resource allocation questions.

Although we are not studying alternative resource allocation methods, it is worth pointing out that if supply and demand is not used, that does not make difficult choices go away. Scarcity means there is not enough to go around. We may decide we do not want to use markets to allocate scarce organs, but we will still need a mechanism to decide whose lives are saved.

This section changes the focus from how supply and demand works to an evaluation of the market system’s solution. The approach is clear: We first consider what an optimal allocation would look like, and then check to see whether the market’s allocation conforms to the optimal solution.

## Finding the Optimal Quantity in a Single Market

To find the optimal solution, we conduct a fanciful analysis. Like the imaginary budget line we used to find income and substitution effects, we work out a thought experiment that actually can never be carried out.

Suppose you had special powers and could allocate resources any way you wanted? Your official title might be *Omniscient, Omnipotent Social Planner*, or OOSP, for short. You are omniscient, or all knowing, so you know every consumer’s desires and every firm’s costs of production. Because you are omnipotent, or all powerful, you can decide how much to produce of each good and service and how it is produced and distributed.

Because this is partial equilibrium analysis, we focus on just one good or service. The question for you, OOSP, is, “How much should be produced of this particular commodity?”

One way for you to answer this question is to measure the total gain obtained by the consumers and producers of the good. When we compute the gain, we subtract the cost of acquiring the product for consumers and, for firms, the costs of production. The plan is to compute the total net gain for different quantities and pick that quantity at which the total gain is maximized.

The notion of net gain, something above the cost that is captured by consumers and firms, is the fundamental idea behind consumers’ and producers’ surplus. *Consumers’ surplus* is the gain from consumption after accounting for the costs of purchasing the product. *Producer’s surplus* is the difference between total revenues and total variable costs. In the long run, it is profit.

We begin with producers’ surplus because it is uncontroversial. We will see that consumers’ surplus is problematic.

## Producers’ Surplus

At any given price, if sellers get that price for all of the units sold, they get a surplus from the sale of each unit except the last one. The sum of these surpluses is the producer’s surplus. The sum of all of the producer’s surpluses in the market is the producers’ surplus, *PS*.

The location of the apostrophe matters. Producer’s surplus is the surplus obtained by one firm. If the focus is on all of the firms, we use producers’ surplus.

*STEP* Open the Excel workbook *CSPS.xls*, read the *Intro* sheet, then go to the *PS* sheet.

The sheet displays an inverse supply curve given by \(P = 35 + 0.52Q_s\). The area of the green triangle is *PS*. To see why, consider the situation when output is 75 units and the price is $74/unit.

The very last unit sold added $74 to total cost (given that we know that the supply curve is the marginal cost curve). Thus, the 75^{th} unit sold yielded no surplus. In general, the marginal unit yields no surplus.

But what about the other units? All of the other units are *inframarginal* units. In other words, these are units below the marginal (last) unit and, in general, the inframarginal units generate surplus. The firm is receiving a price in excess of marginal cost for these units, from 1 to 74, and, therefore, it is reaping a surplus each of those units. We can add them up to get producer’s surplus.

Consider the 50^{th} unit. The marginal cost of the 50^{th} unit is given by \(35 + 0.52 * 50 = \$61\). The firm would have been willing to sell the 50^{th} unit for $61, but it was paid $74 for that 50^{th} unit. So, it made $13 on the 50^{th} unit.

*STEP* Look at cell Q68. It reports the surplus generated by the 50^{th} unit, $13, as we computed above. Look at cell Q28. It reports the surplus generated by the 50^{th} unit, $33.80.

Cell R19 adds the surpluses from all of the inframarginal units. Notice how PS steadily falls from the first to the last unit. The key to PS is that all quantities are sold at the same price, but marginal cost starts low and rises. The firm makes a surplus above *MC* on all output except the last one.

Cell R19 differs from cells B19 and B21 because cell R19 is based on an integer interpretation of output. If output is continuous, then we can compute the *PS* as the area of the triangle created by the horizontal price and the supply curve.

Notice that cell B19 offers another way to understand *PS*. If supply is marginal cost, then the area *under* the marginal cost curve is total variable cost. Because marginal cost is linear, the computation is easy. If *MC* was a curve, we would have to integrate. Total revenue is simply price times quantity. Cell B19 computes \(TR - TVC\), the excess over variable cost, which is the producers’ surplus.

*STEP* If \(Q_s = 95\), what is *PS*? Use the scroll bar in cell C12 to set quantity equal to 95.

At 95 units of units of output, *MC* is $84.40. At that price, the 95^{th} unit has no surplus. But all of the other, inframarginal units generate surplus, adding up to $2,346.50.

*STEP* Explore other quantities and confirm that as output rises, so does producers’ surplus.

## Consumers’ Surplus

The idea is the same. At any given price, if a buyer pays that price for all of the units bought, she gets a surplus from the purchase of each unit except the last one. The sum of these surpluses is the consumer’s surplus. The sum of all of the consumer’s surpluses is the consumers’ surplus, *CS*.

*STEP* Proceed to the *CS* sheet.

Given the inverse demand curve, \(P = 350 - 0.2Q_d\), we can easily compute *CS* for a given quantity as the area of the pink triangle.

At \(Q_d = 95\), the price so consumers will buy 95 units is $160/unit. The last unit purchased provides no surplus, but the inframarginal units generate *CS*. The area under the demand curve, but above the price, is a measure of the net satisfaction enjoyed by consumers.

Consumers’ surplus comes from the fact that consumers would have paid more for each inframarginal unit than the price they actually paid so they get a surplus for each marginal unit.

*STEP* Use the quantity scroll bar to confirm that as output rises, so does consumers’ surplus.

As mentioned earlier, there is a problem with consumers’ surplus. We will finish how OOSP could use *CS* and *PS* before explaining the problem.

## Maximizing *CS* and *PS*

Producers’ surplus is the amount by which the total revenue exceeds variable costs and measures gain for the firm. Consumers’ surplus also measures gain because it is the amount by which the total satisfaction provided by the commodity exceeds the total costs of purchasing the commodity.

Both parties, consumers and producers, gain from trade. This is why a trade is madeboth buyer and seller are better off. When you buy something, you part with some money in exchange for the good or service. If the purchase is voluntary, you must value what you are getting more than what you paid for it or else you would not have bought it. Similarly, the seller values the money you pay more than the good or service or else she would refuse to sell at that price. The gains from voluntary trade are captured in the terms consumers’ and producers’ surplus.

Casting the problem in terms of surplus received by buyers and sellers leads naturally to this question: What is the level of output that maximizes the total surplus? After all, it is clear that as quantity changes the *CS* and *PS* also change.

Thus, OOSP is faced with the following optimization problem: \[\max\limits_{q} CS(q)+PS(q)\] The idea is to maximize the gains from trade for all buyers and sellers. This problem can be solved analytically and numerically. We focus on the latter.

*STEP* Proceed to the *CSandPS* sheet.

This sheet combines the surpluses enjoyed by producers and consumers into a single chart, shown in Figure 17.8.

Understanding this chart is fundamental. We proceed slowly. The vertical dashed line represents the quantity, which OOSP controls and will choose so that *CS* + *PS* is maximized.

There are two prices on the chart, one for the firm and the other for the consumer. The idea is that OOSP uses the quantity to determine the prices needed for firms to be willing to produce the output level and for consumers to want to buy that amount of output.

This is *not* an equilibrium model of supply and demand. OOSP cares only about choosing the optimal output. Price for consumers and firms is used only to compute surplus.

In Figure 17.8 (and on your computer screen), producers receive a price of $84.40 for each of the 95 units, yet consumers pay $160.00 per unit. Remember that OOSP, our benevolent dictator, has magical powers so she can charge one price to consumers and give a different price to producers. By adding the values in cells E18 and B21, we get the value in cell J20. It is highlighted in yellow and maximizing it is the goal.

*STEP* Click on the slider control (over cell C12), to increase output in increments of five units.

As output increases, *CS* and *PS* both rise.

*STEP* Continue clicking on the slider control so that output rises above 125 units.

Now the sum of *CS* and *PS* is falling. That is confusing because the two triangles are getting bigger. But once the price to consumers falls below the price to the firms, we have to pay the difference. This is explained below in more detail. For now, let’s work finding optimal *Q*.

*STEP* Launch Solver and use it to find \(Q \mbox{*}\).

With an empty Solver dialog box, you have to provide the objective (J20) and changing cell (B12). We find that \(CS + PS\) is maximized at \(Q \mbox{*} = 125\) units.

In other words, OOSP should order the manufacture of 125 units of this product, allocating the inputs needed from society’s scarce resource endowment. This level of output maximizes the sum of *CS* and *PS*.

We have seen this number before. In the previous section, we found that the equilibrium solution was \(Q_e=125\) units. This means that the market’s solution is the optimal solution. This is a remarkable result.

No one intended this. No one chose this. No one directed this. Supply and demand established an equilibrium output which answered the question of how much to produce and we now see that it is the same solution we would have chosen if our goal was to maximize consumers’ and producer’s surplus. This is truly amazing.

## Deadweight Loss

If OOSP chooses an output level below 125 and charges a price to consumers based on the inverse demand curve and pays producers a price based on the inverse supply curve, it will generate a smaller value of \(CS + PS\).

How much smaller? The amount of surplus not captured is given by the trapezoid between the consumers’ and producers’ surpluses. This area is called *deadweight loss*, *DWL*. It is a fundamental concept in economics and merits careful attention.

*STEP* Enter 95 in cell B12, then click the button.

Not only do data appear below the button, but the chart has been modified to include a red trapezoid. The area of the trapezoid is displayed in cell D30.

*STEP* Click on cell D26.

The formula is simply the solution of the intersection of the supply and demand curves. We know this quantity is the solution to the problem of maximizing *CS* and *PS*.

*STEP* Click on cell D28.

This seemingly complicated formula is not really that hard. It displays the maximum possible total surplus. Two things are being added, *CS* and *PS*. The first part of the formula is *PS*: 0.5*(((s0_ + s1_ *D26) \(-\) s0_)*D26). It is half the height of the *PS* triangle times the length (or quantity produced). The second part of the formula uses the same area of the triangle formula to compute the *CS*: 0.5*((d0_ - (d0_ \(-\) d1_*D26))*D26).

*STEP* Click on cell D30.

The formula, = D28 \(-\) J20, makes crystal clear that deadweight loss is maximum total surplus minus the sum of *CS* and *PS* at any value of output. In other words, deadweight loss is a measure of the inefficiency of producing the wrong level of output in a particular market. Deadweight loss vaporizes surplus so that it disappears into thin air. Deadweight loss is pure waste.

*STEP* Click on the slider control (over cell C12) to increase output in increments of five units.

As you increase output, note that the deadweight loss falls as the output approaches the optimal quantity. There is no deadweight loss when the output is at 125 because this is the optimal level of output.

Another way to expressing the efficiency in the allocation of resources of the equilibrium solution is to say it has no deadweight loss. That is, no inefficiency in allocating resources.

As *Q* approaches \(Q \mbox{*}\) we reach the maximum possible \(CS + PS\) and *DWL* goes to zero. As *Q* keeps rising, past \(Q > Q \mbox{*}\), we get less total \(CS+PS\) and deadweight loss rises. We get deadweight loss on either side of \(Q \mbox{*}\). The explanation for deadweight loss when \(Q>Q \mbox{*}\) is more complicated. Let’s look at some concrete numbers.

*STEP* Set output above the optimal level, for example, *Q* = 150.

Your screen should look like Figure 17.9. It is true that *CS* and *PS* triangles are large, but with a higher price to firms than consumers, society has to pay for the difference. Once we account for this, the total gain is less than that at \(Q=125\) and we suffer deadweight loss, as shown by the red triangle.

Figure 17.9 shows that it is possible to have sellers receive $113 per unit sold yet have buyers pay only $50 per unit sold, but someone is going to have to make up that $63 per unit difference. The total value of the subsidy, $63/unit times 150 units is $9,340. This amount (rectangle ABCD in Figure 17.9) must be subtracted from the sum of *CS* and *PS*.

When we add everything up, we get a total surplus of $18,900 at \(Q=150\), which is lower than the maximum total surplus. Cell J20 uses an IF statement to get the calculation right. The deadweight loss from producing 150 units is $787.50 (cell D30).

The deadweight loss at \(Q = 150\) is given by the area of the red triangle in Figure 17.9. The geometry is easy. We must subtract a rectangle with height 63 and length 150 from the sum of the pink *CS* and green *PS* triangles. This leaves the red triangle as the *DWL* caused by producing too much output.

There is one optimal output and at that value, deadweight loss is zero. Outputs above and below \(Q \mbox{*}\) produce inefficiency in the allocation of resources because we fail to maximize \(CS + PS\). This is called deadweight loss.

## Price Controls

Price controls are legally mandated limits on prices. A price ceiling sets the highest price at which the good can be legally sold. A price floor does the opposite: The good cannot be sold any lower than the given amount.

To be effective, a price ceiling has to be set below and a price floor has to be set above the equilibrium price.

Most introductory economics students are taught that price ceilings generate shortages and price floors lead to surpluses. For most students, the take-home message is that market forces cannot push the price above the ceiling or below the floor so the market cannot clear and this is why price controls are undesirable.

It turns out that this is not exactly right. Although it is true that ceilings lead to persistent excess demand and floors prevent the market from eliminating excess supply, the real reason behind the unpopularity (among economists) of price controls is the fact that they cause a misallocation of resources.

*STEP* Proceed to the *PriceCeiling* sheet.

Suppose there is a price ceiling on this good at $84.40. At this price, there is a shortage of the good because quantity demanded at $84.40 is 132.8 units (cell B13) while quantity supplied is only 95 (cell B12).

The price cannot be bid up because $84.40 is the highest price at which the good can be legally sold. Thus, with this price ceiling, the output level is 95. We know this is an inefficient result because we know \(Q \mbox{*} = 125\). This is the real reason why this price ceiling is a poor policy, not because it causes a shortage. The price ceiling fails to maximize total surplus.

To be clear, with this price ceiling, too few resources are allocated to the production of this good or service. There will be only 95 units of it produced, not the optimal 125 units. The fact that there is a shortage is true, but it is the misallocation of resources that is the problem.

While the misallocation of resources is easy to see since the quantity is wrong, deadweight loss is more complicated. It depends on the story about the price control and how agents react.

Suppose, for example, that market players are all honest so there is no illegal selling of the good above the maximum price. In other words, producers do not violate the law. Suppose further that the good is allocated via lottery so there are no lines of buyers or resources spent waiting. This means that consumers’ surplus is now a trapezoid instead of a triangle.

*STEP* Click the button.

As shown in Figure 17.10 (and on your screen), a rectangle has been removed from deadweight loss so it is now just the red triangle.

In addition to the usual *CS* triangle in Figure 17.10, consumers enjoy the area of the rectangle computed by multiplying a price of $160 (which is the price consumers are willing to pay for 95 units of the good) minus $84.40 (the price consumers actually pay) times 95 units.

The good news behind this price ceiling with no cheating story is that the deadweight loss is much smaller than in the *CSandPS* sheet with \(Q=95\) because the lucky consumers who can purchase the good do not have to pay $160/unit. The bad news is that there is still a deadweight loss of $1,134. This is a measure of the inefficiency of the price ceiling with no illegal market.

Suppose instead that there are unlawful sales of the product at the illegal market price, $160/unit (this is the most buyers are willing to pay for 95 units). Suppose, in addition, that somehow there are no wasted resources associated with this illegal market. No police investigations, court cases, or any other resources are spent on stopping criminal sales. Then the producers get the rectangle. With this idealized illegal market, the rectangle is transferred from consumers to producers, but the deadweight loss stays the same. The Q&A sheet asks you to demonstrate this.

If, as is almost surely true, illegal selling results in more resources being spent, then the deadweight loss is larger than the red triangle. Illegal activity often leads to violence (think of illegal drugs, which are a market with a price ceiling of zero) and we would subtract that from \(CS+PC\) and thereby increase *DWL*.

Consider two other stories about the price ceiling. A limited set of buyers are given coupons to buy the product. To buy the good (at the legal price), you must have a coupon. If a rationing coupon scheme is used, the sellers of the coupons get the rectangle. The deadweight loss remains the same.

Suppose, finally, that a price ceiling is set and the good is allocated on a first-come-first-serve basis. In other words, buyers have to wait in line. With this story, the resources buyers waste standing in line (or paying others to stand in line for them) must be subtracted from the total surplus. The deadweight loss rises. If the entire rectangle is lost, then the deadweight loss is the same as that in the *CSandPS* sheet when 95 units of output are produced.

Price controls are a popular way to modify market results. Unfortunately, from a resource allocation standpoint, price controls suffer from the fact that they fail to maximize total surplus. It is this property and not that they produce shortages that earn price ceilings criticism. We want the allocation mechanism to give optimal *Q*.

It is confusing that correctly measuring deadweight loss depends on the story, but do not be distracted by the many ways price controls are implemented. The take-home message is that any deviation from \(Q \mbox{*}\) means that the allocation scheme has failed. Deadweight loss, which gives a measure of the inefficiency in monetary units, depends on the specific implementation of the price control, but the fact that it is not zero is evidence that it has failed.

## Caveat Emptor

"Let the buyer beware" is the meaning of the Latin phrase, *caveat emptor*. This idea from contract law is a warning to the buyer that they are responsible for what they are buying. The consumer needs to be careful so they aren’t tricked or end up with a poor quality, unsuitable product.

*Caveat emptor* applies to deadweight loss. On the one hand, deadweight loss is a common way that economists measure inefficiency. It is based on the idea that the maximum total surplus is not attained from a particular output level. But users need to know what they are getting themselves intodeadweight loss has two glaring weaknesses.

The first has to do with our calculation of consumers’ surplus. For technical reasons, restrictive assumptions about the utility function must be imposed. For example, a Cobb-Douglas utility function for individual consumers will not work because it has an income effect. A quasilinear utility function will work (no income effect), but it is unlikely that all consumers have quasilinear utility.

Consumers’ surplus violates the rule that we should not make interpersonal utility comparisons. We are using the demand curve to add up dollar measures of the extra satisfaction that different people get from consuming a product. That is unsound and breaks a basic tenet of modern utility theory.

The second weakness stems from the use of partial equilibrium analysis. We are calculating deadweight loss based on the impact in a single market of a deviation in output from its optimal level. The focus on one market is too limited. If we apply too many or too few resources to the production of one good, we will cause deviations from optimal output for other goods and services. So, the deadweight loss computation based on one market is a lower bound. To get it exactly right, we would have to analyze effects on other markets and do a general equilibrium analysis.

Regarding deadweight loss, it is *caveat emptor*. Remember that deadweight loss measures inefficiency and it is commonly used in applied work, but it is not exactly right. The best way to think of deadweight loss is as an approximation.

Some economists are appalled at the thought of using it, they are usually more theoretically oriented. Economists who do empirical work are more likely to argue that deadweight loss is imperfect, but practically speaking, it is a useful way of measuring inefficiency.

## Optimal Allocation of Resources

This is an important section. It introduced producers’ and consumers’ surpluses, which are key elements in the omnipotent, omniscient social planner’s objective function.

The idea that there is an optimal level of output for each good and service is fundamental. From this idea we get the procedure for evaluating any allocation scheme or government policy: We compare an observed result to the optimal answer.

It is obvious that quantities below the intersection of supply and demand cannot be optimal because both *CS* and *PS* rise as *Q* increases. The situation with quantity above the intersection of supply and demand is more subtle. To get the calculation right, whenever quantity is above the intersection point, we must subtract from the sum of *CS* and *PS* a rectangle that is the difference between prices multiplied by quantity.

The most important and remarkable result from this section is that \(Qe = Q \mbox{*}\). This says that in a properly functioning market, the equilibrium quantity (which is the market system’s answer to society’s resource allocation problem) yields the socially optimal level of output.

Price controls lead to inefficient allocation of resources. The output generated does not match the optimal output. The deadweight loss associated with a price control depends on the story of how the particular implementation of the price control is enforced and responded to by buyers and sellers.

There is no question that deadweight loss is a linchpin of policy analysis. Countless estimates of deadweight loss and cost–benefit studies have been conducted. It is, however, flawed. Measuring consumers’ surplus in value of money terms from a market demand curve in a partial equilibrium setting leaves us on very thin ice. Applications and estimates of deadweight loss should be seen as an approximation to the exact measure of the loss from the misallocation of resources (if such a measure exists).

While deadweight loss is flawed, the notion of a misallocation of resources is not. The idea that there is an optimal solution to society’s resource allocation problem is perfectly valid. So is defining an allocation that deviates from optimal as a misallocation of resources. These are bedrock ideas in microeconomic theory.

This should mark the end of this section, but because there is so much confusion about equilibrium and optimal resource allocation, what follows is an attempt to provide some clarity.

**Equilibrium and Optimal Resource Allocation**

The material below is being repeated for emphasis. The Theory of Consumer Behavior and Theory of the Firm are stepping stones to the \(Q_e = Q \mbox{*}\) result. Let’s put things in perspective and explain why this is so fundamental.

It is absolutely true that philosophers and deep thinkers of the day were baffled by the market system. There was active debate about how and why Europe and, within Europe, England was getting so rich. How could the unplanned, individual decisions of many buyers and sellers produce a pattern, much less a good result? It seemed obvious that a leaderless, fragmented system would produce chaos.

In the previous section, we saw that the equilibrium quantity, \(Q_e\), generated by a properly functioning market is located at the intersection of supply and demand. The market uses a good’s price to send signals to buyers and sellers. Prices above equilibrium are pushed down, whereas prices below equilibrium are pushed up. At the equilibrium solution, the price has no tendency to change and output is also at rest. The equilibrium level of output is the market’s answer to how much of society’s resources will be devoted to producing this particular good.

Our work in this section on consumers’ and producers’ surplus takes a much different perspective on the resource allocation problem. Instead of examining how the market works, we have created a thought experiment, giving an imaginary social planner incredible powers. Given the goal of maximizing total surplus, OOSP would choose an optimal quantity, \(Q \mbox{*}\), that should be produced. If we produce less or more than this socially optimal amount, society would forego surpluses that would make producers and consumers better off. This is called deadweight loss.

If we compare the market’s equilibrium quantity to the socially optimal quantity, we are struck by an amazing result: \(Q_e = Q \mbox{*}\). This critical equivalence means that we do not need a dictator, benevolent or otherwise, to optimally allocate resources. The market, using prices, can settle down to a position of rest where all gains from trade are completely exploited and the sum of producers’ and consumers’ surplus is maximized.

There is no guarantee, however, that \(Q_e=Q \mbox{*}\)there are conditions under which the invisible hand does not lead the market to optimality. We will see examples where the equality does not hold and the market is said to fail.

As you work on this section and this part of the book, do not lose sight of the main point: The market’s ability to generate an equilibrium quantity that is socially optimal is nothing short of amazing and unbelievable. It is equivalent to geese flying a V. A pattern is generated by the interactions of individuals with no awareness or intent to make the pattern.

Consider this hypothetical: we learn that broccoli cures cancer. Would we need a president, prime minister, or king to tell farmers to grow more broccoli? Of course not. Broccoli would fly of the shelves, its price would rocket, and farmers would *automatically* start producing more broccoli.

Analogies from biology are many, but this one might be so shocking and different from anything you have seen before that it will convey why supply and demand is so fascinating to economists.

*STEP* Visit http://tiny.cc/siphonophore to learn about this creature and see it in action.

## Exercises

- From the
*CSandPS*sheet, click the button, then set \(d_0=375\) and use Solver to find the optimal quantity. Take a picture of the cells that contain your answer and paste it in a Word doc. - Click the button. Suppose there was a price ceiling of $84.40. What is the story about price ceilings assumed by the chart and
*DWL*computations on the sheet? - Suppose the government implemented a price support scheme (this is a type of price floor that is used frequently for agricultural products) where they only allowed 95 units to be produced. Cell E16 shows that the market price would be $185. Compute the deadweight loss and explain it.

## References

The epigraph is from the first page of R. W. Houghton, "A Note on the Early History of Consumer’s Surplus," *Economica*, New Series, Vol. 25, No. 97 (February, 1958), pp. 49–57, www.jstor.org/stable/2550693 A French engineer, Jules Dupuit (pronounced doo-pwee) presented the idea of *utilite relative* in 1844, but Alfred Marshall independently rediscovered and popularized the notion of consumer’s surplus.

Almost immediately after Marshall introduced consumers’ surplus, the concept came under attack. It has survived the move from a cardinal to an ordinal perspective on utility and a variety of other criticisms. Economists know that *CS* is built on shaky foundations, but they often use it in practical, policy-oriented, real-world discussions. In a review of the state of *CS*, Abram Bergson concludes, “Despite theoretic criticism, practitioners have continued to apply consumer’s surplus analysis through the years. As some have argued, that must already say something about the usefulness (as well as the use) of such analysis, but just what it says has remained more or less in doubt.” See “A Note on Consumer’s Surplus,” *Journal of Economic Literature*, Vol. 13, No. 1 (March, 1975), pp. 38–44, www.jstor.org/stable/2722212