PART 4
CHAPTER 12
Matching the right buyers with the right sellers creates economic value that is just as real as the value created by the actual production of goods and services.©Reed Saxon/AP Images
LEARNING OBJECTIVES
After reading this chapter, you should be able to:
1. LO1Explain how middlemen add value to market transactions.
2. LO2Use the concept of rational search to find the optimal amount of information market participants should obtain.
3. LO3Define asymmetric information and describe how it influences buying and selling decisions.
4. LO4Discuss how incomplete information leads to statistical discrimination.
5. LO5Understand why people purchase insurance and how the problems of adverse selection and moral hazard affect insurance markets.
Years ago, a naive young economist spent a week in Kashmir on a houseboat on scenic Dal Lake, outside the capital city of Srinagar. Kashmir is renowned for its woodcarvings, and one afternoon a man in a gondola stopped by to show the economist some of his wooden bowls. When the economist expressed interest in one of them, the woodcarver quoted a price of 200 rupees. The economist had lived in that part of Asia long enough to realize that the price was more than the woodcarver expected to get, so he made a counteroffer of 100 rupees.
The woodcarver appeared to take offense, saying that he couldn’t possibly part with the bowl for less than 175 rupees. Suspecting that the woodcarver was merely feigning anger, the young economist held firm. The woodcarver appeared to become even angrier, but quickly retreated to 150 rupees. The economist politely restated his unwillingness to pay more than 100 rupees. The woodcarver then tried 125 rupees, and again the economist replied that 100 was his final offer. Finally, they struck a deal at 100 rupees, and with cash in hand, the woodcarver left in a huff.
Pleased with his purchase, the economist showed it to the houseboat’s owner later that evening. “It’s a lovely bowl,” he agreed, and asked how much the economist had paid for it. The economist told him, expecting praise for his negotiating prowess. The host’s failed attempt at suppressing a laugh was the economist’s first clue that he had paid too much. When asked how much such a bowl would normally sell for, the houseboat owner was reluctant to respond. But the economist pressed him, and the host speculated that the seller had probably hoped for 30 rupees at most.
Adam Smith’s invisible hand theory presumes that buyers are fully informed about the myriad ways in which they might spend their money—what goods and services are available, what prices they sell for, how long they last, how frequently they break down, and so on. But, of course, no one is ever really fully informed about anything. And sometimes, as in the transaction with the woodcarver, people are completely ignorant of even the most basic information. Still, life goes on, and most people muddle through somehow.
Consumers employ a variety of strategies for gathering information, some of which are better than others. They read Consumer Reports, talk to family and friends, visit stores, kick the tires on used cars, and so on. But one of the most important aspects of choosing intelligently without having complete information is having at least some idea of the extent of one’s ignorance. Someone once said that there are two kinds of consumers in the world: those who don’t know what they’re doing and those who don’t know that they don’t know what they’re doing. As in the case of the wooden bowl, the people in the second category are the ones who are most likely to choose foolishly.
Basic economic principles can help you to identify those situations in which additional information is most likely to prove helpful. In this chapter, we will explore what those principles tell us about how much information to acquire and how to make the best use of limited information.
HOW THE MIDDLEMAN ADDS VALUE
One of the most common problems consumers confront is the need to choose among different versions of a product whose many complex features they don’t fully understand. As the following example illustrates, in such cases consumers can sometimes rely on the knowledge of others.
EXAMPLE 12.1Consumer Choice
How should a consumer decide which pair of skis to buy?
You need a new pair of skis, but the technology has changed considerably since you bought your last pair and you don’t know which of the current brands and models would be best for you. Skis R Us has the largest selection, so you go there and ask for advice. The salesperson appears to be well informed; after asking about your experience level and how aggressively you ski, she recommends the Rossignol Experience 77. You buy a pair for $600 and then head back to your apartment and show them to your roommate, who says that you could have bought them on the Internet for only $400. How do you feel about your purchase? Are the different prices charged by the two suppliers related to the services they offer? Were the extra services you got by shopping at Skis R Us worth the extra $200?
Internet retailers can sell for less because their costs are much lower than those of full-service retail stores. Those stores, after all, must hire knowledgeable salespeople, put their merchandise on display, rent space in expensive shopping malls, and so on. Internet retailers and mail-order houses, by contrast, typically employ unskilled telephone clerks, and they store their merchandise in cheap warehouses. If you’re a consumer who doesn’t know which is the right product for you, the extra expense of shopping at a specialty retailer is likely to be a good investment. Spending $600 on the right skis is smarter than spending $400 on the wrong ones.
Many people believe that wholesalers, retailers, and other agents who assist manufacturers in the sale of their products play a far less important role than the one played by those who actually make the products. In this view, the production worker is the ultimate source of economic value added. Sales agents are often disparaged as mere middlemen, parasites on the efforts of others who do the real work.
On a superficial level, this view might seem to be supported by the fact that many people go to great lengths to avoid paying for the services of sales agents. Many manufacturers cater to them by offering consumers a chance to “buy direct” and sidestep the middleman’s commission. But on closer examination, we can see that the economic role of sales agents is essentially the same as that of production workers. Consider this example.
EXAMPLE 12.2The Economic Role of Sales Agents
How does better information affect economic surplus?
Ellis has just inherited a rare Babe Ruth baseball card issued during the great slugger’s rookie year. He’d like to keep the card but has reluctantly decided to sell it to pay some overdue bills. His reservation price for the card is $300, but he is hoping to get significantly more for it. He has two ways of selling it: he can place a classified ad in the local newspaper for $5 or he can list the card on eBay. If he sells the card on eBay, the fee will be 5 percent of the winning bid.
Because Ellis lives in a small town with few potential buyers of rare baseball cards, the local buyer with the highest reservation price is willing to pay $400 at most. If Ellis lists the card on eBay, however, a much larger number of potential buyers will see it. If the two eBay shoppers who are willing to pay the most for Ellis’s card have reservation prices of $900 and $800, respectively, by how much will the total economic surplus be larger if Ellis sells his card on eBay? (For the sake of simplicity, assume that the eBay commission and the classified ad fee equal the respective costs of providing those services.)
In an eBay auction, each bidder reports his or her reservation price for an item. When the auction closes, the bidder with the highest reservation price wins, and the price he or she pays is the reservation price of the second-highest bidder. So in this example, the Babe Ruth baseball card will sell for $800 if Ellis lists it on eBay. Net of the $40 eBay commission, Ellis will receive a payment of $760, or $460 more than his reservation price for the card. Ellis’s economic surplus will thus be $460. The winning bidder’s surplus will be $900 − $800 = $100, so the total surplus from selling the card on eBay will be $560.
If Ellis instead advertises the card in the local newspaper and sells it to the local buyer whose reservation price is $400, then Ellis’s surplus (net of the newspaper’s $5 fee) will be only $95 and the buyer’s surplus will be $0. Thus, total economic surplus will be $560 − $95 = $465 larger if Ellis sells the card on eBay than if he lists it in the local newspaper.
eBay provides a service by making information available to people who can make good use of it. A real increase in economic surplus results when an item ends up in the hands of someone who values it more highly than the person who otherwise would have bought it. That increase is just as valuable as the increase in surplus that results from manufacturing cars, growing corn, or any other productive activity.
RECAP
HOW THE MIDDLEMAN ADDS VALUE
In a world of incomplete information, sales agents and other middlemen add genuine economic value by increasing the extent to which goods and services find their way to the consumers who value them most. For example, when a sales agent causes a good to be purchased by a person who values it by $20,000 more than the person who would have bought it in the absence of a sales agent, that agent augments total economic surplus by $20,000, an achievement on a par with the production of a $20,000 car.
THE OPTIMAL AMOUNT OF INFORMATION
Increasing Opportunity Cost
Without a doubt, having more information is better than having less. But information is generally costly to acquire. In most situations, the value of additional information will decline beyond some point. And because of the Low-Hanging-Fruit Principle, people tend to gather information from the cheapest sources first before turning to more costly ones. Typically, then, the marginal benefit of information will decline, and its marginal cost will rise, as the amount of information gathered increases.
THE COST-BENEFIT TEST
Cost-Benefit
Information gathering is an activity like any other. The Cost-Benefit Principle tells us that a rational consumer will continue to gather information as long as its marginal benefit exceeds its marginal cost. Suppose, for the sake of discussion, that analysts had devised a scale that permits us to measure units of information, as on the horizontal axis of Figure 12.1. If the relevant marginal cost and marginal benefit curves are as shown in the diagram, a rational consumer will acquire I* units of information, the amount for which the marginal benefit of information equals its marginal cost.
FIGURE 12.1 The Optimal Amount of Information.For the marginal cost and benefit curves shown, the optimal amount of information is I*. Beyond that point, information costs more to acquire than it is worth.
Another way to think about Figure 12.1 is that it shows the optimal level of ignorance. When the cost of acquiring information exceeds its benefits, acquiring additional information simply doesn’t pay. If information could be acquired at no cost, decision makers would, of course, be glad to have it. But when the cost of acquiring the information exceeds the gain in value from the decision it will facilitate, people are better off to remain ignorant.
THE FREE-RIDER PROBLEM
Does the invisible hand ensure that the optimal amount of advice will be made available to consumers in the marketplace? The next example suggests one reason why it might not.
The Economic Naturalist 12.1
Why is finding a knowledgeable salesclerk often difficult?
People can choose for themselves whether to bear the extra cost of retail shopping. Those who value advice and convenience can pay slightly higher prices, while those who know what they want can buy for less from a mail-order house. True or false: It follows that private incentives lead to the optimal amount of retail service.
Why are there so few knowledgeable salesclerks?
The market would provide the optimal level of retail service except for one practical problem, namely, that consumers can make use of the services offered by retail stores without paying for them. After benefiting from the advice of informed salespersons and after inspecting the merchandise, the consumer can return home and buy the same item from an Internet retailer or mail-order house. Not all consumers do so, of course. But the fact that customers can benefit from the information provided by retail stores without paying for it is an example of the free-rider problem, an incentive problem that results in too little of a good or service being produced. Because retail stores have difficulty recovering the cost of providing information, private incentives are likely to yield less than the socially optimal level of retail service. So the statement above is false.
The Economic Naturalist 12.2
Why did Rivergate Books, the last bookstore in Lambertville, New Jersey, go out of business?
Small independent bookstores often manage to survive competition from large chains like Barnes & Noble by offering more personalized service. Janet Holbrooke, the proprietor of Rivergate Books, followed this strategy successfully for more than a decade before closing her doors in 1999. What finally led her to quit?
Mrs. Holbrooke, a retired English teacher, said that many of her customers had experimented with Barnes & Noble when the large chain opened an outlet nearby, but that most had remained loyal because they valued the more personal service her clerks were able to provide. Customers also were drawn in by special events such as readings and book signings by authors. But during one of these events, Mrs. Holbrooke saw that her store’s days were numbered. She had invited Gerald Stern, a National Book Award–winning poet, to do a reading, but despite a good turnout for the event, few attendees purchased books. When she overheard a woman in the book-signing line say that she’d brought her book to the event after having purchased it from Amazon.com, Mrs. Holbrook realized it was a losing battle.
Why are so many independent booksellers going out of business?
CONCEPT CHECK 12.1
Apart from its possible contribution to free-rider problems, how is increased access to the Internet likely to affect total economic surplus?
TWO GUIDELINES FOR RATIONAL SEARCH
Cost-Benefit
In practice, of course, the exact value of additional information is difficult to know, so the amount of time and effort one should invest in acquiring it is not always obvious. But as the following examples suggest, the Cost-Benefit Principle provides a strong conceptual framework for thinking about this problem.
EXAMPLE 12.3Searching for an Apartment
Should a person living in Paris, Texas, spend more or less time searching for an apartment than someone living in Paris, France?
Suppose that rents for one-bedroom apartments in Paris, Texas, vary between $300 and $500 per month, with an average rent of $400 per month. Rents for similar one-bedroom apartments in Paris, France, vary between $2,000 and $3,000 per month, with an average rent of $2,500. In which city should a rational person expect to spend a longer time searching for an apartment?
In both cities, visiting additional apartments entails a cost, largely the opportunity cost of one’s time. In both cities, the more apartments someone visits, the more likely it is that he or she will find one near the lower end of the rent distribution. But because rents are higher and are spread over a broader range in Paris, France, the expected saving from further time spent searching will be greater there than in Paris, Texas. A rational person will expect to spend more time searching for an apartment in France.
This example illustrates the principle that spending additional search time is more likely to be worthwhile for expensive items than for cheap ones. For example, one should spend more time searching for a good price on a diamond engagement ring than for a good price on a stone made of cubic zirconium; more time searching for a low fare to Sydney, Australia, than for a low fare to Sidney, New York; and more time searching for a car than for a bicycle. By extension, hiring an agent—someone who can assist with a search—is more likely to be a good investment in searching for something expensive than for something cheap. For example, people typically engage real estate agents to help them find a house, but they seldom hire agents to help them buy a gallon of milk.
EXAMPLE 12.4The Cost of Searching
Who should expect to search longer for a good price on a used piano?
Both Tom and Tim are shopping for a used upright piano. To examine a piano listed in the classified ads, they must travel to the home of the piano’s current owner. If Tom has a car and Tim does not and both are rational, which one should expect to examine fewer pianos before making his purchase?
The benefits of examining an additional piano are the same in both cases, namely, a better chance of finding a good instrument for a low price. But because it is more costly for Tim to examine pianos, he should expect to examine fewer of them than Tom.
The preceding example makes the point that when searching becomes more costly, we should expect to do less of it. And as a result, the prices we expect to pay will be higher when the cost of a search is higher.
THE GAMBLE INHERENT IN SEARCH
Suppose you are in the market for a one-bedroom apartment and have found one that rents for $400 per month. Should you rent it or search further in hopes of finding a cheaper apartment? Even in a large market with many vacant apartments, there is no guarantee that searching further will turn up a cheaper or better apartment. Searching further entails a cost, which might outweigh the gain. In general, someone who engages in further search must accept certain costs in return for unknown benefits. Thus, further search invariably carries an element of risk.
In thinking about whether to take any gamble, a helpful first step is to compute its expected value—the average amount you would win (or lose) if you played that gamble an infinite number of times. To calculate the expected value of a gamble with more than one outcome, we first multiply each outcome by its corresponding probability of occurring, and then add. For example, suppose you win $1 if a coin flip comes up heads and lose $1 if it comes up tails. Since 1/2 is the probability of heads (and also the probability of tails), the expected value of this gamble is (1/2)($1) + (1/2)(−$1) = 0. A gamble with an expected value of zero is called a fair gamble. If you played this gamble a large number of times, you wouldn’t expect to make money, but you also wouldn’t expect to lose money.
A better-than-fair gamble is one with a positive expected value. (For instance, a coin flip in which you win $2 for heads and lose $1 for tails is a better-than-fair gamble.) A risk-neutral person is someone who would accept any gamble that is fair or better. A risk-averse person is someone who would refuse to take any fair gamble.
CONCEPT CHECK 12.2
Consider a gamble in which you win $4 if you flip a fair coin and it comes up heads and lose $2 if it comes up tails. What is the expected value of this gamble? Would a risk-neutral person accept it?
In the next example, we apply these concepts to the decision of whether to search further for an apartment.
EXAMPLE 12.5The Gamble in the Search
Should you search further for an apartment?
You have arrived in San Francisco for a one-month summer visit and are searching for a one-bedroom sublet for the month. There are only two kinds of one- bedroom apartments in the neighborhood in which you wish to live, identical in every respect except that one rents for $400 and the other for $360. Of the vacant apartments in this neighborhood, 80 percent are of the first type and 20 percent are of the second type. The only way you can discover the rent for a vacant apartment is to visit it in person. The first apartment you visit is one that rents for $400. If you are risk-neutral and your opportunity cost of visiting an additional apartment is $6, should you visit another apartment or rent the one you’ve found?
If you visit one more apartment, you have a 20 percent chance of it being one that rents for $360 and an 80 percent chance of it being one that rents for $400. If the former, you’ll save $40 in rent, but if the latter, you’ll face the same rent as before. Since the cost of a visit is $6, visiting another apartment is a gamble with a 20 percent chance to win $40 − $6 = $34 and an 80 percent chance of losing $6 (which means “winning” −$6). The expected value of this gamble is thus (0.20)($34) + (0.80)(−$6) = $2. Visiting another apartment is a better-than-fair gamble, and since you are risk-neutral, you should take it.
CONCEPT CHECK 12.3
Refer to Example 12.5. Suppose you visit another apartment and discover it, too, is one that rents for $400. If you are risk-neutral, should you visit a third apartment?
THE COMMITMENT PROBLEM WHEN SEARCH IS COSTLY
When most people search for an apartment, they want a place to live not for just a month, but for a year or more. Most landlords, for their part, are also looking for long-term tenants. Similarly, few people accept a full-time job in their chosen field unless they expect to hold the job for several years. Firms, too, generally prefer employees who will stay for extended periods. Finally, when most people search for mates, they are looking for someone with whom to settle down.
Because in all these cases search is costly, examining every possible option will never make sense. Apartment hunters don’t visit every vacant apartment, nor do landlords interview every possible tenant. Job seekers don’t visit every employer, nor do employers interview every job seeker. And not even the most determined searcher can manage to date every eligible mate. In these and other cases, people are rational to end their searches, even though they know a more attractive option surely exists out there somewhere.
But herein lies a difficulty. What happens when, by chance, a more attractive option comes along after the search has ceased? Few people would rent an apartment if they thought the landlord would kick them out the moment another tenant came along who was willing to pay higher rent. Few landlords would be willing to rent to a tenant if they expected her to move out the moment she discovers a cheaper apartment. Employers, job seekers, and people who are looking for mates would have similar reservations about entering relationships that could be terminated once a better option happened to come along.
This potential difficulty in maintaining stable matches between partners in ongoing relationships would not arise in a world of perfect information. In such a world, everyone would end up in the best possible relationship, so no one would be tempted to renege. But when information is costly and the search must be limited, there will always be the potential for existing relationships to dissolve.
In most contexts, people solve this problem not by conducting an exhaustive search (which is usually impossible, in any event) but by committing themselves to remain in a relationship once a mutual agreement has been reached to terminate the search. Thus, landlords and tenants sign a lease that binds them to one another for a specified period, usually one year. Employees and firms enter into employment contracts, either formal or informal, under which each promises to honor his obligations to the other, except under extreme circumstances. And in most countries a marriage contract penalizes those who abandon their spouses. Entering into such commitments limits the freedom to pursue one’s own interests. Yet most people freely accept such restrictions because they know the alternative is failure to solve the search problem.
RECAP
THE OPTIMAL AMOUNT OF INFORMATION
Additional information creates value, but it’s also costly to acquire. A rational consumer will continue to acquire information until its marginal benefit equals its marginal cost. Beyond that point, it’s rational to remain uninformed.
Markets for information do not always function perfectly. Free-rider problems often hinder retailers’ efforts to provide information to consumers.
Search inevitably entails an element of risk because costs must be incurred without any assurance that search will prove fruitful. A rational consumer can minimize this risk by concentrating search efforts on goods for which the variation in price or quality is relatively high and on those for which the cost of search is relatively low.
ASYMMETRIC INFORMATION
One of the most common information problems occurs when the participants in a potential exchange are not equally well informed about the product or service that’s offered for sale. For instance, the owner of a used car may know that the car is in excellent mechanical condition, but potential buyers cannot know that merely by inspecting it or taking it for a test drive. Economists use the term asymmetric information to describe situations in which buyers and sellers are not equally well informed about the characteristics of products or services. In these situations, sellers are typically much better informed than buyers, but sometimes the reverse will be true.
The problem of asymmetric information can easily prevent exchanges that would benefit both parties. Here is a classic example.
EXAMPLE 12.6Asymmetric Information
Will Jane sell her car to Tom?
Jane’s 2014 Mazda Miata has 70,000 miles on the odometer, but most of these are highway miles driven during weekend trips to see her boyfriend in Toronto. (Highway driving causes less wear and tear on a car than city driving.) Moreover, Jane has maintained the car precisely according to the manufacturer’s specifications. In short, she knows her car to be in excellent condition. Because she is about to start graduate school in Boston, however, Jane wants to sell the car. On average, 2014 Miatas sell for a price of $8,000, but because Jane knows her car to be in excellent condition, her reservation price for it is $10,000.
Tom wants to buy a used Miata. He would be willing to pay $13,000 for one that is in excellent condition but only $9,000 for one that is not in excellent condition. Tom has no way of telling whether Jane’s Miata is in excellent condition. (He could hire a mechanic to examine the car, but doing so is expensive, and many problems cannot be detected even by a mechanic.) Will Tom buy Jane’s car? Is this outcome efficient?
Because Jane’s car looks no different from other 2014 Miatas, Tom will not pay $10,000 for it. After all, for only $8,000, he can buy some other 2014 Miata that’s in just as good condition, as far as he can tell. Tom therefore will buy someone else’s Miata, and Jane’s will go unsold. This outcome is not efficient. If Tom had bought Jane’s Miata for, say, $11,000, his surplus would have been $2,000 and Jane’s another $1,000. Instead, Tom ends up buying a Miata that is in average condition (or worse), and his surplus is only $1,000. Jane gets no economic surplus at all.
THE LEMONS MODEL
We can’t be sure, of course, that the Miata Tom ends up buying will be in worse condition than Jane’s—since someone might have a car in perfect condition that must be sold even if the owner cannot get what it is really worth. Even so, the economic incentives created by asymmetric information suggest that most used cars that are put up for sale will be of lower-than-average quality. One reason is that people who mistreat their cars, or whose cars were never very good to begin with, are more likely than others to want to sell them. Buyers know from experience that cars for sale on the used car market are more likely to be “lemons” than cars that are not for sale. This realization causes them to lower their reservation prices for a used car.
But that’s not the end of the story. Once used car prices have fallen, the owners of cars that are in good condition have an even stronger incentive to hold onto them. That causes the average quality of the cars offered for sale on the used car market to decline still further. Berkeley economist George Akerlof, a Nobel laureate, was the first to explain the logic behind this downward spiral.1 Economists use the term lemons model to describe Akerlof’s explanation of how asymmetric information affects the average quality of the used goods offered for sale.
The next example suggests that the lemons model has important practical implications for consumer choice.
EXAMPLE 12.7The Lemons Model in Action
Should you buy your aunt’s car?
You want to buy a used Honda Accord. Your Aunt Germaine buys a new car every four years, and she has a four-year-old Accord that she’s about to trade in. You believe her report that the car is in good condition, and she’s willing to sell it to you for $10,000, which is the current blue book value for four-year-old Accords. (The blue book value of a car is the average price for which cars of that age and model sell in the used car market.) Should you buy your aunt’s Honda?
Akerlof’s lemons model tells us that cars for sale in the used car market will be of lower average quality than cars of the same vintage that are not for sale. If you believe your aunt’s claim that her car is in good condition, then being able to buy it for its blue book value is definitely a good deal for you, since the blue book price is the equilibrium price for a car that is of lower quality than your aunt’s.
The following two examples illustrate the conditions under which asymmetric information about product quality results in a market in which only lemons are offered for sale.
EXAMPLE 12.8The Naive Buyer (Part 1)
How much will a naive buyer pay for a used car?
Consider a world with only two kinds of cars: good ones and lemons. An owner knows with certainty which type of car she has, but potential buyers cannot distinguish between the two types. Ten percent of all new cars produced are lemons. Good used cars are worth $10,000 to their owners, but lemons are worth only $6,000. Consider a naive consumer who believes that the used cars currently for sale have the same quality distribution as new cars (i.e., 90 percent good, 10 percent lemons). If this consumer is risk-neutral, how much would he be willing to pay for a used car?
Buying a car of unknown quality is a gamble, but a risk-neutral buyer would be willing to take the gamble provided it is fair. If the buyer can’t tell the difference between a good car and a lemon, the probability that he will end up with a lemon is simply the proportion of lemons among the cars from which he chooses. The buyer believes he has a 90 percent chance of getting a good car and a 10 percent chance of getting a lemon. Given the prices he is willing to pay for the two types of car, his expected value of the car he buys will thus be 0.90($10,000) + 0.10($6,000) = $9,600. And since he is risk-neutral, that is his reservation price for a used car.
CONCEPT CHECK 12.4
How would your answer to the question posed in Example 12.8 differ if the proportion of new cars that are lemons had been 20 percent?
EXAMPLE 12.9The Naive Buyer (Part 2)
Who will sell a used car for what the naive buyer is willing to pay?
Continuing with the previous example: If you were the owner of a good used car, what would it be worth to you? Would you sell it to a naive buyer? What if you owned a lemon?
Since you know your car is good, it is worth $10,000 to you, by assumption. But a naive buyer would be willing to pay only $9,600, so neither you nor any other owner of a good car would be willing to sell to that buyer. If you had a lemon, of course, you’d be happy to sell it to a naive buyer, since the $9,600 the buyer is willing to pay is $3,600 more than the lemon would be worth to you. So the only used cars for sale will be lemons. In time, buyers will revise their naively optimistic beliefs about the quality of the cars for sale on the used car market. In the end, all used cars will sell for a price of $6,000, and all will be lemons.
In practice, of course, the mere fact that a car is for sale does not guarantee that it is a lemon because the owner of a good car will sometimes be forced to sell it, even at a price that does not reflect its condition. The logic of the lemons model explains this owner’s frustration. The first thing sellers in this situation want a prospective buyer to know is the reason they are selling their cars. For example, classified ads often announce, “Just had a baby, must sell my 2015 Corvette” or “Transferred to Germany, must sell my 2016 Toyota Camry.” Any time you pay the blue book price for a used car that is for sale for some reason unrelated to its condition, you are beating the market.
THE CREDIBILITY PROBLEM IN TRADING
Why can’t someone with a high-quality used car simply tell the buyer about the car’s condition? The difficulty is that buyers’ and sellers’ interests tend to conflict. Sellers of used cars, for example, have an economic incentive to overstate the quality of their products. Buyers, for their part, have an incentive to understate the amount they are willing to pay for used cars and other products (in the hope of bargaining for a lower price). Potential employees may be tempted to overstate their qualifications for a job. And people searching for mates have been known to engage in deception.
That isn’t to say that most people consciously misrepresent the truth in communicating with their potential trading partners. But people do tend to interpret ambiguous information in ways that promote their own interests. Thus, 92 percent of factory employees surveyed in one study rated themselves as more productive than the average factory worker. Psychologists call this phenomenon the “Lake Wobegon effect,” after Garrison Keillor’s mythical Minnesota homestead, where “all the children are above average.”
Why do new cars lose a significant fraction of their value as soon as they are driven from the showroom?
Notwithstanding the natural tendency to exaggerate, the parties to a potential exchange can often gain if they can find some means to communicate their knowledge truthfully. In general, however, mere statements of relevant information will not suffice. People have long since learned to discount the used car salesman’s inflated claims about the cars he’s trying to unload. But as the next example illustrates, though communication between potential adversaries may be difficult, it’s not impossible.
EXAMPLE 12.10Credible Signals
How can a used car seller signal high-quality credibly?
Jane knows her Miata to be in excellent condition, and Tom would be willing to pay considerably more than her reservation price if he could be confident of getting such a car. What kind of signal about the car’s quality would Tom find credible?
Again, the potential conflict between Tom’s and Jane’s interests suggests that mere statements about the car’s quality may not be persuasive. But suppose Jane offers a warranty, under which she agrees to remedy any defects the car develops over the next six months. Jane can afford to extend such an offer because she knows her car is unlikely to need expensive repairs. In contrast, the person who knows his car has a cracked engine block would never extend such an offer. The warranty is a credible signal that the car is in good condition. It enables Tom to buy the car with confidence, to both his and Jane’s benefit.
The Costly-to-Fake Principle
The preceding examples illustrate the costly-to-fake principle, which holds that if parties whose interests potentially conflict are to communicate credibly with one another, the signals they send must be costly or difficult to fake. If the seller of a defective car could offer an extensive warranty just as easily as the seller of a good car, a warranty offer would communicate nothing about the car’s quality. But warranties entail costs that are significantly higher for defective cars than for good cars—hence their credibility as a signal of product quality.
To the extent that sellers have an incentive to portray a product in the most flattering light possible, their interests conflict with those of buyers, who want the most accurate assessment of product quality possible. Note that in the following example, the costly-to-fake principle applies to a producer’s statement about the quality of a product.
The Economic Naturalist 12.3
Why do firms insert the phrase “As advertised on TV” when they advertise their products in magazines and social media?
Company A sponsors an expensive national television advertising campaign on behalf of its surround sound speakers, claiming it has the clearest sound and the best repair record of any surround sound speakers in the market. Company B makes similar claims in a sales brochure but does not advertise its product on television. If you had no additional information to go on, which company’s claim would you find more credible? Why do you suppose Company A mentions its TV ads when it advertises its surround sound speakers in print and social media?
Why should buyers care whether a product is advertised on TV?
Accustomed as we are to discounting advertisers’ inflated claims, the information given might seem to provide no real basis for a choice between the two products. On closer examination, however, we see that a company’s decision to advertise its product on national television constitutes a credible signal about the product’s quality. The cost of a national television campaign can run well into the millions of dollars, a sum a company would be foolish to spend on an inferior product.
For example, Budweiser paid Arnold Schwartzenegger $3 million to appear in its 2014 Superbowl ad for Bud Light, and the going rate for 30-second ad slots for that year’s game was $4 million. National TV ads can attract the potential buyers’ attention and persuade a small fraction of them to try a product. But these huge investments pay off only if the resulting initial sales generate other new business—either repeat sales to people who tried the product and liked it or sales to others who heard about the product from a friend.
Because ads cannot persuade buyers that a bad product is a good one, a company that spends millions of dollars advertising a bad product is wasting its money. An expensive national advertising campaign is therefore a credible signal that the producer thinks its product is a good one. Of course, the ads don’t guarantee that a product is a winner, but in an uncertain world, they provide one more piece of information. Note, however, that the relevant information lies in the expenditure on the advertising campaign, not in what the ads themselves say.
These observations may explain why some companies mention their television ads in their print ads. Advertisers understand the costly-to-fake principle and hope that consumers will understand it as well.
As the next Economic Naturalist example illustrates, the costly-to-fake principle is also well known to many employers.
The Economic Naturalist 12.4
Why do many companies care so much about elite educational credentials?
Microsoft is looking for a hardworking, smart person for an entry-level managerial position in a new technical products division. Two candidates, Cooper and Duncan, seem alike in every respect but one, Cooper, graduated with the highest honors from MIT, while Duncan graduated with a C+ average from Somerville College. Whom should Microsoft hire?
Why do some employers care so much about elite degrees?
If you want to persuade prospective employers that you are both hardworking and intelligent, there is perhaps no more credible signal than to have graduated with distinction from a highly selective educational institution. Most people would like potential employers to think of them as hardworking and intelligent. But unless you actually have both those qualities, graduating with the highest honors from a school like MIT will be extremely difficult. The fact that Duncan graduated from a much less selective institution and earned only a C+ average is not proof positive that he is not diligent and talented, but companies are forced to play the percentages. In this case, the odds strongly favor Cooper.
Conspicuous Consumption as a Signal of Ability
Cost-Benefit
Some individuals of high ability are not highly paid. (Remember the best elementary school teacher you ever had.) And some people—such as the multibillionaire investor Warren Buffet—earn a lot, yet spend very little. But such cases run counter to general tendencies. In competitive markets, the people with the most ability tend to receive the highest salaries. And as suggested by the Cost-Benefit Principle, the more someone earns, the more he or she is likely to spend on high-quality goods and services. As the following example suggests, these tendencies often lead us to infer a person’s ability from the amount and quality of the goods he consumes.
The Economic Naturalist 12.5
Why do many clients seem to prefer lawyers who wear expensive suits?
You’ve been unjustly accused of a serious crime and are looking for an attorney. Your choice is between two lawyers who appear identical in all respects except for the things they buy. One of them wears a cheap polyester suit and arrives at the courthouse in a 10-year-old rust-eaten Dodge Caliber. The other wears an impeccably tailored suit and drives a new BMW 750i. If this were the only information available to you at the time you chose, which lawyer would you hire?
If you were on trial for a serious crime, which lawyer would you hire?
The correlation between salary and the abilities buyers value most is particularly strong in the legal profession. A lawyer whose clients usually prevail in court will be much more in demand than one whose clients generally lose, and their fees will reflect the difference. The fact that one of the lawyers consumes much more than the other doesn’t prove that he is the better lawyer, but if that is the only information you have, you can ill afford to ignore it.
If the less able lawyer loses business because of the suits he wears and the car he drives, why doesn’t he simply buy better suits and a more expensive car? His choice is between saving for retirement or spending more on his car and clothing. In one sense, he cannot afford to buy a more expensive car, but in another sense, he cannot afford not to. If his current car is discouraging potential clients from hiring him, buying a better one may simply be a prudent investment. But because all lawyers have an incentive to make such investments, their effects tend to be mutually offsetting.
When all is said and done, the things people consume will continue to convey relevant information about their respective ability levels. The costly-to-fake principle tells us that the BMW 750i is an effective signal precisely because the lawyer of low ability cannot afford one, no matter how little he saves for retirement. Yet from a social perspective, the resulting spending pattern is inefficient, for the same reason that other positional arms races are inefficient. (We discussed positional arms races in Chapter 11, Externalities, Property Rights, and the Environment.) Society would be better off if everyone spent less and saved more for retirement.
The problem of conspicuous consumption as an ability signal doesn’t arise with equal force in every environment. In small towns, where people tend to know one another well, a lawyer who tries to impress people by spending beyond her means is likely to succeed only in demonstrating how foolish she is. Thus, the wardrobe a professional person “needs” in towns like Dubuque, Iowa, or Athens, Ohio, costs less than half as much as the wardrobe the same person would need in Manhattan or Los Angeles.
RECAP
ASYMMETRIC INFORMATION
· Asymmetric information describes situations in which not all parties to a potential exchange are equally well informed. In the typical case, the seller of a product will know more about its quality than the potential buyers. Such asymmetries often stand in the way of mutually beneficial exchange in the markets for high-quality goods because buyers’ inability to identify high quality makes them unwilling to pay a commensurate price.
· Information asymmetries and other communication problems between potential exchange partners can often be solved through the use of signals that are costly or difficult to fake. Product warranties are such a signal because the seller of a low-quality product would find them too costly to offer.
STATISTICAL DISCRIMINATION
In a competitive market with perfect information, the buyer of a service would pay the seller’s cost of providing the service. In many markets, however, the seller does not know the exact cost of serving each individual buyer.
In such cases, the missing information has an economic value. If the seller can come up with even a rough estimate of the missing information, she can improve her position. As Economic Naturalist 12.6 illustrates, firms often do so by imputing characteristics to individuals on the basis of the groups to which they belong.
The Economic Naturalist 12.6
Why do males under 25 years of age pay more than other drivers for auto insurance?
Gerald is 23 years old and is an extremely careful and competent driver. He has never had an accident, or even a moving traffic violation. His twin sister Geraldine has had two accidents, one of them serious, in the last three years and has accumulated three speeding tickets during that same period. Why does Gerald pay $1,600 per year for auto insurance, while Geraldine pays only $800?
The expected cost to an insurance company of insuring any given driver depends on the probability that the driver will be involved in an accident. No one knows what that probability is for any given driver, but insurance companies can estimate rather precisely the proportion of drivers in specific groups who will be involved in an accident in any given year. Males under 25 are much more likely than older males and females of any age to become involved in auto accidents. Gerald pays more than his sister because even those males under 25 who have never had an accident are more likely to have one than females the same age who have had several accidents.
Why do male teens pay so much more for auto insurance?
Of course, females who have had two accidents and accumulated several tickets in the last three years are more likely to have an accident than a female with a spotless driving record. The insurance company knows that and has increased Geraldine’s premium accordingly. Yet it is still less than her brother’s premium. That doesn’t mean that Gerald is in fact more likely to have an accident than Geraldine. Indeed, given the twins’ respective driving skills, Geraldine clearly poses the higher risk. But because insurance companies lack such detailed information, they are forced to set rates according to the information they possess.
To remain in business, an insurance company must collect enough money from premiums to cover the cost of the claims it pays out, plus whatever administrative expenses it incurs. Consider an insurance company that charges lower rates for young males with clean driving records than for females with blemished ones. Given that the former group is more likely to have accidents than the latter, the company cannot break even unless it charges females more, and males less, than the respective costs of insuring them. But if it does so, rival insurance companies will see cash on the table: They can offer females slightly lower rates and lure them away from the first company. The first company will end up with only young male policyholders and thus will suffer an economic loss at the low rates it charges. That is why, in equilibrium, young males with clean driving records pay higher insurance rates than young females with blemished records.
The insurance industry’s policy of charging high rates to young male drivers is an example of statistical discrimination. Other examples include the common practice of paying higher salaries to people with college degrees than to people without them and the policy of favoring college applicants with high SAT scores. Statistical discrimination occurs whenever people or products are judged on the basis of the groups to which they belong.
Competition promotes statistical discrimination, even though everyone knows that the characteristics of specific individuals can differ markedly from those of the group to which they belong. For example, insurance companies know perfectly well that some young males are careful and competent drivers. But unless they can identify which males are the better drivers, competitive pressure forces them to act on their knowledge that, as a group, young males are more likely than others to generate insurance claims.
Similarly, employers know that many people with only a high school diploma are more productive than the average college graduate. But because employers usually cannot tell in advance who those people are, competitive pressure leads them to offer higher wages to college graduates, who are more productive, on average, than high school graduates. Universities, too, realize that many applicants with low SAT scores will earn higher grades than applicants with high scores. But if two applicants look equally promising except for their SAT scores, competition forces universities to favor the applicant with higher scores since, on average, that applicant will perform better than the other.
Statistical discrimination is the result of observable differences in group characteristics, not the cause of those differences. Young males, for example, do not generate more insurance claims because of statistical discrimination. Rather, statistical discrimination occurs because insurance companies know that young males generate more claims. Nor does statistical discrimination cause young males to pay insurance rates that are high in relation to the claims they generate. Among any group of young male drivers, some are careful and competent, and others are not. Statistical discrimination means the more able males will pay high rates relative to the volume of claims they generate, but it also means the less able male drivers will pay low rates relative to the claims they generate. On average, the group’s rates will be appropriate to the claims its members generate.
Still, these observations do little to ease the frustration of the young male who knows himself to be a careful and competent driver, or the high school graduate who knows herself to be a highly productive employee. Competitive forces provide firms an incentive to identify such individuals and treat them more favorably whenever practical. When firms succeed in this effort, however, they have often discovered some other relevant information on group differences. For example, many insurance companies offer lower rates to young males who belong to the National Honor Society or make the dean’s list at school. Members of those groups generate fewer claims, on average, than other young males. But even these groups include risky drivers, and the fact that companies offer discounts to their members means that all other young males must pay higher rates.
DISAPPEARING POLITICAL DISCOURSE
An intriguing illustration of statistical discrimination arises when a politician decides what to say about controversial public issues. Politicians have an interest in supporting the positions they genuinely believe in, but they also have an interest in winning reelection. As Economic Naturalist 12.7 and 12.8 illustrate, the two motives often conflict, especially when a politician’s statements about one subject convey information about her beliefs on other subjects.
The Economic Naturalist 12.7
Why do opponents of the death penalty often remain silent?
Quite apart from the question of whether execution of convicted criminals is morally legitimate, there are important practical arguments against capital punishment. For one thing, it is extremely expensive relative to the alternative of life without parole. Execution is costly because of judicial safeguards against execution of innocent persons. In each capital case prosecuted in the United States, these safeguards consume thousands of person-hours from attorneys and other officers of the court, at a cost that runs well into the millions of dollars.2 Such efforts notwithstanding, the record is replete with examples of executed persons who are later shown to be innocent. Another argument against capital punishment is that many statistical studies find that it does not deter people from committing capital crimes. Though many political leaders in both parties find these and other arguments against capital punishment compelling, few politicians voice their opposition to capital punishment publicly. Why not?
A possible answer to this puzzle is suggested by the theory of statistical discrimination. Voters in both parties are concerned about crime and want to elect politicians who take the problem seriously. Suppose there are two kinds of politicians: some who in their heart of hearts take the crime issue seriously and others who merely pay lip service to it. Suppose also that voters classify politicians in a second way: those who publicly favor the death penalty or remain silent and those who publicly oppose it. Some politicians will oppose the death penalty for the reasons just discussed, but others will oppose it because they are simply reluctant to punish criminals—perhaps because they believe that crime is ultimately more society’s fault than the criminal’s. (Politicians in the latter category are the ones voters think of as being “not serious about crime”; they are the ones many voters want to get rid of.) These two possible motives for opposing the death penalty suggest that the proportion of death penalty opponents who take the crime issue seriously, in the public’s view, will be somewhat smaller than the corresponding proportion among proponents of the death penalty. For the sake of discussion, imagine that 95 percent of politicians who favor the death penalty and only 80 percent of politicians who oppose the death penalty are “serious about crime.”
If you are a voter who cares about crime, how will your views about a politician be affected by hearing that he opposes the death penalty? If you knew nothing about that politician to begin with, your best guess on hearing his opposition to the death penalty would be that there is an 80 percent chance that he is serious about crime. Had he instead voiced support for the death penalty, your best guess would be that there is a 95 percent chance that he is serious about crime. And since voters are looking for politicians who are serious about crime, the mere act of speaking out against the death penalty will entail a small loss of political support even for those politicians who are extremely serious about crime.
Knowing this tendency on the part of voters, some politicians who are only marginally opposed to the death penalty may prefer to keep their views to themselves. As a result, the composition of the group that speaks out publicly against the death penalty will change slightly so that it is more heavily weighted with people reluctant to punish criminals in any way. Suppose, for example, that the proportion of death penalty opponents who are serious about crime falls from 80 to 60 percent. Now the political cost of speaking out against the death penalty rises, leading still more opponents to remain silent. Once the dust settles, very few opponents of capital punishment will risk stating their views publicly. In their desire to convince voters that they are tough on crime, some may even become outspoken proponents of the death penalty. In the end, public discourse will strongly favor capital punishment. But that is no reason to conclude that most leaders—or even most voters—genuinely favor it.
The economist Glenn Loury was the first to call attention to the phenomenon described in the preceding example. We call it the problem of disappearing political discourse. Once you understand it, you will begin to notice examples not just in the political sphere but in everyday discourse as well.
The Economic Naturalist 12.8
Why do proponents of legalized drugs remain silent?
That addictive drugs like heroin, cocaine, and methamphetamines cause enormous harm is not a matter of dispute. The clear intent of laws that ban commerce in these drugs is to prevent that harm. But the laws also entail costs. By making the drugs illegal, they substantially increase their price, leading many addicts to commit crimes to pay for drugs. The high incomes of many illicit drug dealers also divert people from legitimate careers and result in turf battles that often have devastating consequences for both participants and bystanders. If these drugs were legal, drug-related crime would vanish completely. Drug use would also rise, how significantly we do not know. In short, it’s at least conceivable that legalizing addictive drugs might be sound public policy. Why, then, do virtually no politicians publicly favor such a policy?
Many politicians may simply believe that legalizing drugs is a bad idea. Theoretically, legalization could lead to such a steep rise in drug consumption that the cost of the policy might far outweigh its benefits. This concern, however, is not supported by experience in countries such as England and the Netherlands, which have tried limited forms of legalization. A second explanation is that politicians who favor legalization are reluctant to speak out for fear that others will misinterpret them. Suppose that some people favor legalization based on careful analysis of the costs and benefits, while other proponents are merely crazy. If the proportion of crazies is higher among supporters than among opponents of legalization, someone who speaks out in favor of legalization may cause those who do not know her to increase their estimate of the likelihood she is crazy. This possibility deters some proponents from speaking out, which raises the proportion of crazies among the remaining public supporters of legalization—and so on in a downward spiral, until most of the remaining public supporters really are crazy.
The disappearing political discourse problem helps to explain why the United States had difficulty reestablishing normal diplomatic relations with China, which were severed in the wake of the communist revolution. One could oppose communist expansionism and yet still favor normalized relations with China on the grounds that war is less likely when antagonists communicate openly. In the Cold War environment, however, American politicians were under enormous pressure to demonstrate their steadfast opposition to communism at every opportunity. Fearing that support for the normalization of relations with China would be misinterpreted as a sign of softness toward communism, many supporters of the policy remained silent. Not until Richard Nixon—whose anticommunist credentials no one could question—was elected president were diplomatic relations with China finally reopened.
The problem of disappearing discourse also helps explain the impoverished state of public debate on issues such as the reform of Social Security, Medicare, and other entitlement programs.
RECAP
STATISTICAL DISCRIMINATION
· Buyers and sellers respond to incomplete information by attempting to judge the qualities of products and people on the basis of the groups to which they belong. For example, auto insurance companies charge young male drivers higher rates because they know young men are frequently involved in accidents.
· Because people’s beliefs about different things tend to be correlated, knowing what someone believes about one issue provides at least some clue to his or her beliefs about others. The problem of disappearing political discourse arises because politicians are often reluctant to speak out about some issues publicly for fear that doing so may suggest that they hold unpopular beliefs about related issues.
INSURANCE
Hurricanes, fires, car accidents, and illnesses are largely unforeseeable events. Since most people are risk averse, they often purchase insurance to help guard against the possibility of facing the enormous expenses associated with such events.
To see how this works, consider David. Most of the time, he is healthy, but he knows there’s a chance he could face a serious illness. In particular, suppose there’s a 75 percent chance David won’t have any health care expenses all year, but there’s a 25 percent chance he’ll face a serious illness and will have to pay $16,000 in medical bills. David’s annual health care expenses are thus a gamble with an expected value of (0.75)($0) + (0.25)($16,000) = $4,000. This means that if David is risk averse, he’d rather pay $4,000 each year than face a 25 percent chance of paying $16,000. In fact, David might be willing to pay substantially more than $4,000 to avoid this risk.
This is exactly where insurance steps in. Suppose David is willing to pay $4,500 a year to avoid the risk of having to pay $16,000 in medical bills. An insurance company can charge David $4,500 a year in exchange for the promise of paying his medical expenses should he face a serious illness. This arrangement makes David better off because he no longer faces a 25 percent chance of having to pay $16,000 in medical bills should he fall ill, and it’s beneficial for the insurance company, since it gets $4,500 a year from David, but only expects to pay out $4,000. This extra $500 paid to the insurance company can go towards the cost of administering the insurance plan. And, unlike David, the insurance company doesn’t mind the risk associated with David’s health care costs because the company has lots of other customers just like him, some of whom will have good years and others of whom will have bad years, so that, on average, the company pays only $4,000 a year for each customer.
Despite these obvious benefits, insurance markets often do not function as well as we would like. As we discuss next, the problem of incomplete information again plays a role.
ADVERSE SELECTION
One of the most pernicious problems in insurance markets relates to asymmetric information. To remain in business, an insurance company must set premiums that are high enough to pay for the expenses it has promised to cover. Thus, for example, health insurance companies determine premiums based on the expected medical expenses of the average person. The problem is that individuals often have more information about the likelihood that they will need insurance than do the companies providing the insurance. For example, David’s health insurance company might not know that he has a family history of diabetes. This means that the premiums set by insurance companies are a better deal for high-risk individuals than for low-risk individuals, and as a result, high-risk individuals are more likely to buy insurance than low-risk individuals. This pattern is known as adverse selection. Because of it, insurance companies raise their premiums, which makes buying insurance even less attractive to low-risk individuals, which raises still further the average risk level of those who remain insured. In some cases, only those individuals faced with extreme risks may continue to find insurance an attractive purchase.
MORAL HAZARD
Moral hazard is another problem that makes buying insurance less attractive for the average person. This problem refers to the fact that some people take fewer precautions when they know they are insured. Someone whose car is insured, for example, may take less care to prevent it from being damaged or stolen. Driving cautiously and searching for safe parking spaces require effort, after all, and if the losses from failing to engage in these precautions are covered by insurance, some people will become less vigilant.
Insurance companies help many of their potential clients soften the consequences of problems like moral hazard and adverse selection by offering policies with deductible provisions. Under the terms of an automobile collision insurance policy with, say, a $1,000 deductible provision, the insurance company covers only those collision repair costs in excess of $1,000. For example, if you have an accident in which $3,000 in damage occurs, the insurance company covers only $2,000, and you pay the remaining $1,000.
How does the availability of such policies mitigate the negative effects of adverse selection and moral hazard? Because policies with high deductibles are cheaper for insurance companies to provide, they sell for lower prices. The lower prices represent a much better bargain, however, for those drivers who are least likely to file insurance claims since those drivers are least likely to incur any uncovered repair costs. Policies with deductible provisions also confront careless drivers with more of the extra costs for which they are responsible, giving them additional incentives to take precautions.
These policies benefit insurance buyers in another way. Because the holder of a policy with a deductible provision will not file a claim at all if the damage to his car in an accident is less than the deductible threshold, insurance companies require fewer resources to process and investigate claims, savings that get passed along in the form of lower premiums.
THE PROBLEM WITH HEALTH CARE PROVISION THROUGH PRIVATE INSURANCE
It is troubling, but perhaps not surprising, that access to medical care is extremely limited in many of the world’s poorest nations. After all, citizens of those nations lack enough income to buy adequate food, shelter, and many other basic goods and services. What is surprising, however, is that almost 50 million Americans had no health coverage of any kind when President Barack Obama took office in 2009. In a country as wealthy as the United States, why was this the case?
The answer to that question is rooted in the fact that the United States was almost alone among the world’s nations in its reliance on unregulated private insurance markets to orchestrate the delivery of health care to its citizens. Unregulated private insurance markets are a deeply flawed mechanism for providing access to health care because of the adverse-selection problem. As discussed above, if insurance companies set their rates based on the expected medical expenses of the average person, its policy will seem like a bargain to potential customers who know themselves to be in bad health. At the same time, its policies will seem overpriced to those who know themselves to be in excellent health. The upshot is that a disproportionate share of the customers it attracts will have below-average health status, which means its initial premiums will be too low to cover its costs. To stay in business, it will have to raise its rates. But then potential customers in good health will find its policies even less attractive. A downward spiral often ensues, with the end result that insurance becomes unaffordable for most people.
As discussed in Chapter 10, An Introduction to Behavioral Economics, reliance upon unregulated private health insurance markets in the United States was essentially a historical accident, a consequence of the fact that many labor unions managed to negotiate employer-provided health insurance as part of their compensation packages during the rapidly growing economy of the immediate post–WWII years. Under government policy, employer expenditures for health insurance were nontaxable. Employer-provided insurance was thus much cheaper for employees than private insurance purchased individually with income on which they had already been taxed. That incentive induced nonunion employers to join their union counterparts in offering employer-provided health insurance. And as long as health care spending was a fairly small share of total income, coverage was broad and the system functioned reasonably well.
An important policy detail was that eligibility for the tax exemption was conditional on insurance being made available to all employees irrespective of preexisting medical conditions. Given the high cost of treating individuals with chronic medical problems, private insurance companies are generally reluctant to issue policies to people with serious health problems. But by covering large groups of employees, only a small percentage of whom would be likely to have serious health problems during any year, insurance companies could issue these policies without taking unacceptable risks. Indeed, the large new employer-provided insurance market was sufficiently lucrative that most insurance companies were eager to participate in it.
Although the employer-provided group insurance approach helped keep the adverse- selection problem at bay for many years, this approach began to unravel as medical costs continued to rise relative to all other goods and services. With health insurance premiums taking a bigger and bigger bite out of workers’ paychecks and heightened competition forcing companies to look for new ways to cut costs, some began offering higher wages in lieu of employer-provided health coverage. Younger, healthier workers—for whom medical expenses are normally small—found these offers increasingly tempting.
Parents who didn’t buy health insurance for their families were once viewed as irresponsible, but this stigma lost some of its sting as the number of uninsured grew. As more and more people took jobs without health coverage, going without insurance became more socially acceptable. Making matters worse was the changing composition of the pool of the insured. As more healthy families took jobs without coverage, those left tended to be sicker and more costly to treat, forcing premiums to rise still more rapidly. In short, because of adverse selection, our health insurance system was caught in a long-term death spiral.
THE AFFORDABLE CARE ACT OF 2010
Passed by Congress and signed into law by President Obama in March 2010, the Affordable Care Act was the government’s first serious attempt to halt that death spiral. It contained three main provisions, each one of which was essential for reform to succeed. First, it required insurance companies to offer coverage to everyone on roughly equal terms, irrespective of preexisting medical conditions. Without this provision, the economic imperative of every private insurance company would have been to take every step possible to deny coverage to anyone expected to incur significant medical expenses. Any insurance system that couldn’t cover those who most need care would clearly be unacceptable.
Because insurance companies cannot cover their costs if they insure only the least healthy people, it was also necessary for the Affordable Care Act to include a mandate requiring everyone to buy health insurance. Without a mandate, healthy individuals would face strong incentives to go without health insurance until they got sick, since they would then be able to buy affordable insurance from companies that were forbidden to charge high rates based on preexisting conditions.
The third major feature of the Affordable Care Act was to provide for subsidies to low-income families. You can’t require people to buy insurance if they can’t afford it. With health care costs already high and rapidly rising, it was essential to include some provision to ease the burden on those who are unable to pay.
The act contained numerous other provisions, many of which were designed to slow the rate of health care costs by requiring more streamlined medical record keeping and supporting research on the questions of which treatments were most effective. But the essence of the act lies in its three main provisions—nondiscrimination on the basis of preexisting conditions, the mandate, and subsidies for low-income families. Without any one of these provisions, the decline in the percentage of Americans with health coverage would have surely continued.
In tax legislation enacted in late 2017, Congress included a measure repealing the Affordable Care Act’s mandate. This step promises to reduce participation rates of healthy young people, with adverse effects on the costs of serving those who remain insured. Without further legislative action to maintain the diversity of the insured pool, the health care industry’s long-term death spiral may be poised to resume.
RECAP
INSURANCE
Risk-averse people will purchase insurance to guard themselves against the risk of unanticipated and costly events like hurricanes, car accidents and illness. Adverse selection drives up the cost of insurance because individuals who are the most expensive for companies to insure have the greatest incentive to purchase insurance. Moral hazard also drives up insurance premiums because people have less incentive to protect items that have been insured against theft or damage. In the United States, the problem of adverse selection has left large numbers of Americans without health care coverage and has led to ongoing efforts to reform our heath care system.
Mounting insurance premiums have caused many people in good health to do without health coverage, resulting in higher premiums for those who remain insured. The Affordable Care Act of 2010 was enacted in an attempt to remedy market failures that exist in attempts to provide health care access through unregulated private insurance contracts. In the wake of Congressional repeal of the ACA’s mandate provision in late 2017, the historical decline in the proportion of Americans with health insurance appears poised to resume.
SUMMARY
· Retailers and other sales agents are important sources of information. To the extent that they enable consumers to find the right products and services, they add economic value. In that sense they are no less productive than the workers who manufacture goods or perform services directly. Unfortunately, the free-rider problem often prevents firms from offering useful product information. (LO1)
· Virtually every market exchange takes place on the basis of less-than-complete information. More information is beneficial both to buyers and to sellers, but information is costly to acquire. The rational individual therefore acquires information only up to the point at which its marginal benefit equals its marginal cost. Beyond that point, one is rational to remain ignorant. (LO2)
· Several principles govern the rational search for information. Searching more intensively makes sense when the cost of a search is low, when quality is highly variable, or when prices vary widely. Further search is always a gamble. A risk-neutral person will search whenever the expected gains outweigh the expected costs. A rational search will always terminate before all possible options have been investigated. Thus, in a search for a partner in an ongoing bilateral relationship, there is always the possibility that a better partner will turn up after the search is over. In most contexts, people deal with this problem by entering into contracts that commit them to their partners once they have mutually agreed to terminate the search. (LO2)
· Many potentially beneficial transactions are prevented from taking place by asymmetric information—the fact that one party lacks information that the other has. For example, the owner of a used car knows whether it is in good condition, but potential buyers do not. Even though a buyer may be willing to pay more for a good car than the owner of such a car would require, the fact that the buyer cannot be sure she is getting a good car often discourages the sale. More generally, asymmetric information often prevents sellers from supplying the same quality level that consumers would be willing to pay for. (LO3)
· Both buyers and sellers often can gain by finding ways of communicating what they know to one another. But because of the potential conflict between the interests of buyers and sellers, mere statements about the relevant information may not be credible. For a signal between potential trading partners to be credible, it must be costly to fake. For instance, the owner of a high-quality used car can credibly signal the car’s quality by offering a warranty—an offer that the seller of a low-quality car could not afford to make. (LO3)
· Firms and consumers often try to estimate missing information by making use of what they know about the groups to which people or things belong. For example, insurance firms estimate the risk of insuring individual young male drivers on the basis of the accident rates for young males as a group. This practice is known as statistical discrimination. Statistical discrimination helps to explain the phenomenon of disappearing political discourse, which occurs when opponents of a practice such as the death penalty remain silent when the issue is discussed publicly. (LO4)
· Risk-averse people will purchase insurance to guard themselves against the risk of unanticipated and costly events like hurricanes, car accidents and illness. Adverse selection drives up the cost of insurance because individuals who are the most expensive for companies to insure have the greatest incentive to purchase insurance. In some cases, adverse selection can lead insurance to become so costly that few choose to purchase it. Moral hazard also drives up insurance premiums because people have less incentive to protect items that have been insured against theft or damage. In the United States, the problem of adverse selection has left large numbers of Americans without health care coverage and has led to ongoing efforts to reform our heath care system. (LO5)
KEY TERMS
adverse selection
asymmetric information
better-than-fair gamble
costly-to-fake principle
disappearing political discourse
expected value of a gamble
fair gamble
free-rider problem
lemons model
moral hazard
risk-averse person
risk-neutral person
statistical discrimination
REVIEW QUESTIONS
1. 1.Explain why a gallery owner who sells a painting might actually create more economic surplus than the artist who painted it. (LO1)
2. 2.Can it be rational for a consumer to buy a Chevrolet without having first taken test drives in competing models built by Ford, Chrysler, Honda, Toyota, and others? (LO2)
3. 3.Explain why used cars offered for sale are different, on average, from used cars not offered for sale. (LO3)
4. 4.Explain why the used car market would be likely to function more efficiently in a community in which moral norms of honesty are strong than in a community in which such norms are weak. (LO3)
5. 5.Explain why banks are generally more willing to lend money to someone who is employed than to someone who is unemployed. (LO4)
6. 6.Disability insurance pays a portion of an individual’s wages should a non-work related injury or illness prevent that person from being able to work. Explain why disability insurance premiums might be lower if everyone were required to purchase disability insurance. (LO5)
PROBLEMS
1. 1.Carlos is risk-neutral and has an ancient farmhouse with great character for sale in Slaterville Springs. His reservation price for the house is $130,000. The only possible local buyer is Whitney, whose reservation price for the house is $150,000. The only other houses on the market are modern ranch houses that sell for $125,000, which is exactly equal to each potential buyer’s reservation price for such a house. Suppose that if Carlos does not hire a realtor, Whitney will learn from her neighbor that Carlos’s house is for sale and will buy it for $140,000. However, if Carlos hires a realtor, he knows that the realtor will put him in touch with an enthusiast for old farmhouses who is willing to pay up to $300,000 for the house. Carlos also knows that if he and this person negotiate, they will agree on a price of $250,000. If realtors charge a commission of 5 percent of the selling price and all realtors have opportunity costs of $2,000 for negotiating a sale, will Carlos hire a realtor? If so, how will total economic surplus be affected? (LO1)
2. 2.Ann and Barbara are computer programmers in Nashville who are planning to move to Seattle. Each owns a house that has just been appraised for $100,000. But whereas Ann’s house is one of hundreds of highly similar houses in a large, well-known suburban development, Barbara’s is the only one that was built from her architect’s design. Who will benefit more by hiring a realtor to assist in selling her house, Ann or Barbara? (LO1)
3. 3.Brokers who sell stocks over the Internet can serve many more customers than those who transact business by mail or over the phone. How will the expansion of Internet access affect the average incomes of stockbrokers who continue to do business in the traditional way? (LO2)
4. 4.Whose income do you predict will be more affected by the expansion of Internet access: (LO2)
a. Stockbrokers or lawyers?
b. Doctors or pharmacists?
c. Bookstore owners or the owners of galleries that sell original oil paintings?
5. 5.Fred, a retired accountant, and Jim, a government manager, are 63-year-old identical twins who collect antique pottery. Each has an annual income of $100,000 (Fred’s from a pension, Jim’s from salary). One buys most of his pottery at local auctions, and the other buys most of his from a local dealer. Which brother is more likely to buy at an auction, and does he pay more or less than his brother who buys from the local dealer? (LO2)
6. 6.How will growing Internet access affect the number of film actors and musicians who have active fan clubs? (LO2)
7. 7.Consumers know that some fraction x of all new cars produced and sold in the market are defective. The defective ones cannot be identified except by those who own them. Cars do not depreciate with use. Consumers are risk-neutral and value nondefective cars at $10,000 each. New cars sell for $5,000 and used ones for $2,500. What is the fraction x? (LO3)
8. 8.State whether the following are true or false, and briefly explain why: (LO3)
a. Companies spend billions of dollars advertising their products on network TV primarily because the texts of their advertisements persuade consumers that the advertised products are of high quality.
b. You may not get the optimal level of advice from a retail shop when you go in to buy a lamp for your bike because of the free-rider problem.
c. If you need a lawyer, and all your legal expenses are covered by insurance, you should always choose the best-dressed lawyer with the most expensive car and the most ostentatiously furnished office.
d. The benefit of searching for a spouse is affected by the size of the community you live in.
9. 9.For each pair of occupations listed, identify the one for which the kind of car a person drives is more likely to be a good indication of how good she is at her job. (LO3)
a. Elementary school teacher, real estate agent
b. Dentist, municipal government administrator
c. Engineer in the private sector, engineer in the military
10. 10.Female heads of state (e.g., Israel’s Golda Meir, India’s Indira Gandhi, Britain’s Margaret Thatcher) have often been described as more bellicose in foreign policy matters than the average male head of state. Using Loury’s theory of disappearing discourse, suggest an explanation for this pattern. (LO4)
11. 11.Julie knows there’s a 2 percent chance that her house will be destroyed by fire next year, which would require $250,000 to rebuild (and if her house is not destroyed, she won’t have to pay anything). If Julie is risk-averse, will she pay $5,000 a year for an insurance policy that covers the full cost of rebuilding her house should it be destroyed by fire? Might she pay $5,200? What about $5,400? (LO5)
12. 12.Suppose people cannot tell for sure whether they will fall ill in any given year. High-risk people correctly perceive that their chance of falling ill in a year is 30 percent, and low-risk people correctly perceive that their chance of falling ill is 10 percent. Both high-risk people and low-risk people have to pay $10,000 in medical expenses if they fall ill, and nothing if they remain healthy. (LO5)
a. What are the expected annual medical expenses of a high-risk person?
b. What are the expected annual medical expenses of a low-risk person?
c. Suppose insurance companies cannot tell whether someone is high-risk or low-risk. They only know that half of all people are high-risk, and half of all people are low-risk. So, they offer a health insurance policy that costs $2,000 per year in exchange for covering all of a person’s medical expenses should they fall ill. If high risk-people and low-risk people are risk-neutral, then who will purchase this insurance policy? Will the insurance company be able to stay in business if it continues to charge $2,000 per year? Briefly explain.
ANSWERS TO CONCEPT CHECKS
1. 11.1An Internet search is a cheap way to acquire information about many goods and services, so the effect of increased Internet access will be a downward shift in the supply curve of information. In equilibrium, people will acquire more information, and the goods and services they buy will more closely resemble those they would have chosen in an ideal world with perfect information. These effects will cause total economic surplus to grow. Some of these gains, however, might be offset if the Internet makes the free-rider problem more serious. (LO2)
2. 11.2The probability of getting heads is 0.5, the same as the probability of getting tails. Thus, the expected value of this gamble is (0.5)($4) + (0.5)(−$2) = $1. Since the gamble is better than fair, a risk-neutral person would accept it. (LO2)
3. 11.3Since you still have a 20 percent chance of finding a cheaper apartment if you make another visit, the expected outcome of the gamble is again $2, and you should search again. The bad outcome of any previous search is a sunk cost and should not influence your decision about whether to search again. (LO2)
4. 11.4The expected value of a new car will now be (0.8)($10,000) + (0.2)($6,000) = $9,200. Any risk-neutral consumer who believed that the quality distribution of used cars for sale was the same as the quality distribution of new cars off the assembly line would be willing to pay $9,200 for a used car. (LO3)
1George Akerlof, “The Market for Lemons,” Quarterly Journal of Economics 84 (1970), pp. 488–500.
2See Philip J. Cook and Donna B. Slawson, “The Costs of Processing Murder Cases in North Carolina,” The Sanford Institute of Public Policy, Duke University, Durham, NC, 1993.