OCRd and published on the internet in the public interest by Sanjeev Sabhlok on 6 December 2017. This may be read along with Richard Ebeling’s Capitalism and Information Asymmetry.

Trust for Hire: Voluntary Remedies for Quality and Safety - DANIEL B. KLEIN

ContentsA COMPARATIVE-INSTITUTIONS APPROACH...............................................................................................................2

TRUST IN MARKETS: SOME BASIC IDEAS.................................................................................................................2

Pointed Knowledge and the Information Path....................................................................................................3

Pointed Knowledge Can Obviate the Role of Trust, or Provide a Warrant for It.................................................3

Trusters and Promisers........................................................................................................................................4

Product Characteristics.......................................................................................................................................4

Avoiding the Need for Trust: Increments, Hostages, and Prepaid Plans.............................................................4

Chatting and Getting Face-to-Face Clues.............................................................................................................6

Extended Dealings: Time Wounds All Heels........................................................................................................7

INFORMATION ON QUALITY AND CHARACTER: WHO WILL PROVIDE IT? °.............................................................7

Do-It-Yourself Information: Informal Channels of Information Sharing..............................................................8

Fee for Information.............................................................................................................................................8

Information, No Charge: Self-Disclosure by Promisers........................................................................................9

A Classification of Independent Knower Organizations.....................................................................................12

“Prologue,” by Consumer Health Services (CHS)...............................................................................................12

Ideas for Further Research on Knower Organizations.......................................................................................12

Ways of Apprehending Untrustworthiness.......................................................................................................13

MARKET FORMS AND EXTENDED DEALINGS.........................................................................................................14

The Free-rider Properties of Extended Dealings....................................................................................................15

Trustworthy Promisers Cultivate Extended Dealings.........................................................................................15

The Umbrella of the Brand Name......................................................................................................................15

Dealers Make for Extended Dealings.................................................................................................................16

Reputational Nexus and the Middleman...........................................................................................................17

The Middleman Also Acts as Knower................................................................................................................18

CONCLUDING REMARKS........................................................................................................................................19

1

The market’s myriad decentralized actions do not themselves ensure adequate safety. Centralized controls of various sorts are needed. These have been instituted in the form of regulations, constraints, information programs, licensing and certification . . .

— Jerome Rothenberg (1993, 172)

POLICY COMMENTATORS OFTEN CITE CONSUMER IGNORANCE AS GROUNDS FOR government regulation of product quality and safety. Yet policymakers and scholars themselves do not always show deep knowledge of how voluntary institutions—market-based or otherwise—respond to problems of consumer ignorance. Very little has been written in general terms about how real-world institutions cope with consumer ignorance.1

“The political economy of information remedies has not been explored by the literature to date,” laments the Federal Trade Commission economist Pauline Ippolito (1986, 25, 15). “[Whether an arrangement] that depends more on public enforcement is better than one that depends totally on private reputations is an empirical question ... on which we have virtually no evidence.” Despite the paucity of literature on voluntary reputational mechanisms, during the past century regulators have continued to find the free market unacceptable, like a jury that neglects to hear the defense of the accused.2 Not much has changed since Thomas Moore wrote in 1961 (p. 102), “In general, little is known concerning what the market would require in the absence of licensing.”

A COMPARATIVE-INSTITUTIONS APPROACHA comparative-institutions approach to quality and safety issues tells us that it is not sufficient to identify the imperfections of a laissez-faire system. “The merits or demerits” of such an institutional arrangement, writes Paul Samuelson (1964, 83), “cannot be settled by an appeal to abstract reasoning or principles.” To really compare laissez-faire to alternative institutional arrangements we have to explore their imperfections as well, and some-how weigh the arrangements against each other. A rounded case for laissez-faire, therefore, would need to explore the record of dirigiste policies. There is ample research on the consequences of occupational licensing, the Food and Drug Administration, and so on; most of it is critical to one degree or another.3 4 5

My goal is not to argue that laissez-faire will achieve perfection in quality and safety, nor is it to offer the rounded case for laissez-faire. I happen to believe that most, perhaps all, quality and safety regulations should be eliminated, that a system based on common-law principles and on voluntary institutions and customs would be an

1 Economists have devised models that show how reputation al incentives improve quality (Klein & Leffler 1981; Shapiro 1983; Allen 1984; Biglaiser 1993), game theorists have crafted scenarios which give rise to reputational incentives (Heal 1976; Milgrom, North & Weingast 1990; Kandori 1992; Klein 1992), and scholars have discussed reputational mechanisms in par-ticular industries (Mitchell and Maloney 1989; P. Ippolito and Mathios 1990; Haas-Wilson 1990; Firth 1990; Carter and Manaster 1990; Paba 1991; R. Ippolito 1992).2 Lee Benham (1980, 17) writes: “Almost all licensed occupations have claimed they will successfully cope with undesirable market failures. Frequently there has been little or no evidence in support of the proposition that such externalities exist or that the proposed solution will improve the situation. The absence of systematic evidence in support of such claims has, however, never appeared to act as a deterrent.”3 Studies finding that consumer costs and practitioners’ fees and incomes rise with licensing stringency include Shepard (1978), White (1978), and Pazderka & Muzondo (1983).4 For dental practice in Canada, Evans (1980, 242) summarizes the survey finding of the Saskatchewan Dental Plan Report: “[A] very high proportion-80 to 90 per cent—of the work of a general dentist (time, procedures, or billings) can be performed by a high school graduate with twenty months’ post-secondary training. Quality standards for these procedures match or exceed those of general dentists. These facts are no longer challenged by organized dentistry.” Evans suggests that a dental care system in Canada that made use of paraprofessionals and somehow eliminated the superfluity of dentists could reduce costs by 30 to 40 percent. Lave and Lave (1970, 257-260) suggest that in pediatrics in the United States, 80 percent of the practice could be done by less qualified staff.5 Easy access to basic information and service is prevented by regulations on practicing, diagnosing, and referring (as well as by many laws on business operation). Dental hygienists, for example, in most states, cannot clean teeth without a dentist “over the shoulder,” and they can make neither elementary diagnoses nor referrals. Similar “turf’ restrictions apply to many regulated occupations. Opportune knowledge is also prevented by restrictions on advertising and labeling.

2

improvement on our current regulatory arrangement. My goal is to contribute to the case for laissez-faire in these areas by discussing the wide variety of ways in which voluntary processes mitigate trust problems.

TRUST IN MARKETS: SOME BASIC IDEASThe growth of scientific knowledge of mental and physical health and disease is working a revolution in this matter, undermining the idea that each individual is the best judge of his own desires.

— John Maurice Clark (1936, 99)

Pointed Knowledge and the Information Pathwoman with a broken air conditioner economizes on attention; she pierces the problem with questions that point her down a path of remedy. She seeks not general knowledge of condensation and thermodynamics, but rather pointed knowledge about how to restore comfort in her home. The air conditioner blows warm air. She consults her yellow-pages for “Air Conditioner—Repairs and Service”; she consults that listing for servicemen in her area; she calls one and asks him what’s wrong. He replies that most likely the air conditioner is either out of freon or has a leak in the system. It is costly to advance to the next step, and if servicemen are required to get a license the cost will be greater than if they are not.6

In seeking pointed knowledge, the woman traces an information path from ignorance to a remedy for her troubles. Her path is something like the call number on a library book, where each successive character in the number vastly narrows the search. Consumers do not always trace the best information paths, but there is great value in being able to discover opportunities and formulate alternatives.

Restrictions on consumer choice are often defended on the grounds that consumers do not know enough to choose competently. How much do I know about, say, a medical ailment chosen at random? Indeed, not much. But the importance of the idea of the information path, and the shortcoming of the foregoing quotation from J.M. Clark, is that it is not our random ignorance that governs our fate, but our pointed, purposive knowledge. When in doubt, we ask another question.

Basic information is usually all one needs.7 Just as the biggest library, if it is in disorder, may not be as useful as a small but well-arranged library, a system of easy access to basic knowledge may serve better than a system of limited access to extraordinary knowledge.’ Because our particular problem calls for but one volume in a vast library — indeed, to a good extent the very notions of quality and safety should be individualized (Higgs 1994)—it is important that we can efficiently sift through opportune information when formulating a decision.

Pointed Knowledge Can Obviate the Role of Trust, or Provide a Warrant for ItPointed knowledge helps one get high quality in two ways. In inquiring about her airconditioner problem, the woman may come to apprehend the technical measures required; she may even find that she is able to fix it herself. If she can trace a good information path, information is no longer “asymmetric,” and she circumvents the need for trust.

But technical information is not the only kind of pointed knowledge. Information that does nothing, in the narrow sense, to resolve information asymmetry may nonetheless lead to precisely the pointed knowledge that the woman needs. The woman may seek to discover not what is wrong with her air conditioner, but simply who is an honest serviceman. Her pointed knowledge may not provide a direct demonstration of quality, but rather an as-surance of quality, just as in rhetoric our warrant for an idea may be, not a direct demonstration of its validity, but

6 Feldstein (1993, 329) remarks: “In a for-profit organization or in an HMO comprised of a large number of physicians, the quality of care practised by any one physician in the group affects the reputation, and hence the incomes, of the other physicians.”7 “Prepaid plans ... have not been criticized for producing lower quality medical outcomes than their fee-for-service competitors” (Feldstein 1993, 329).

3

its endorsement by a trusted authority. Knowledge is the product of the individual, made from information and emerging from interpretation and judgement in the realm of human purpose.

Diego Gambetta (1988, ix) aptly describes trust as “the belief on which cooperation is predicted.” If the woman cannot hope to gain the technical insight that permits her to circumvent trust, she instead searches for a warrant for trust. The information she obtains may tell her not about how her air conditioner operates, but how the serviceman operates. Trust itself becomes a valued input to the consumer-producer relationship, and those who can provide it will tend to prosper.

Trusters and PromisersIn most of the market transactions that are subject to quality and safety regulation, the issue of trust is one sided. Typically the issue is whether the consumer should trust the seller. In the absence of more effective terminology, I use the terms “truster” and “promiser.”

The truster (characterized throughout this paper as female) may be the customer who patronizes a doctor, an auto mechanic, an accountant, a lawyer, or a securities broker. She may be a landlord, a creditor, an employee at a potentially hazardous manufacturing plant, or an employer enrolling her work-force in a health-care organization. She must decide whether to entrust her resources to the promiser (characterized as male). The truster is wary and attempts to assess whether the promiser indeed has the incentive or inclination to honor her trust. If her trust is honored she gains from the interaction, but if instead the promiser cheats her, she loses. Wariness may prevent interaction altogether, interaction that sometimes would have turned out well for both parties.

Sometimes the truster is considering the purchase of a standardized product. She asks herself, Is the artificial sweetener really safe? Will the aluminum siding really last 30 years? The answers to her questions have been determined in advance by businessmen, who, in offering a product, act as promisers to a class of trusters.

Product CharacteristicsSometimes we are unable to determine even after the fact whether the promiser cheated or honored; these characteristics of service are called credence characteristics (Darby & Karni 1973). As trusters we are wary in the presence of credence characteristics because often the dealings involve the advisor-reaper tension: the promiser acts as advisor of proper service, but also as the deliverer of that service and, hence, reaper of corresponding returns. We worry that the doctor, lawyer, or mechanic will advise excessive and inappropriate service because he reaps the profit. The economist Robert Evans (1980, 226) remarks that this tension “seems to be the essence of the professional role.” In this case, the content of the promise is, not specified deliverables (whether goods or services), but to deliver whatever service a responsible and conscientious promiser would deem to be in the truster’s interest. The advisor-reaper tension may be alleviated in not-for-profit organizations (Hansmann 1980).

Less vexing characteristics are divided into search characteristics, which can be ascertained prior to purchase, such as the body design of a new car, and experience characteristics, which are determined by experiencing the service, such as the styling of a haircut (Nelson 1974).

A trust problem resembles a “prisoner’s dilemma,” and when play of the game is frequent, cooperation may come forth. But even in infrequent dealings the actors may invent means of resolving their problems. Sometimes the need for trust can be avoided altogether. Sometimes it can be reconstituted and put into another form that is more manageable. Players do not like prison-ers’ dilemmas; they or others have incentives to create alternative games with different rules and happier outcomes.

Avoiding the Need for Trust: Increments, Hostages, and Prepaid PlansSome search and experience characteristics can be conveyed by advertisements, displays, free samples, and try-out periods. These are ways of incrementalizing the trading relationship. In some states dental offices may give

4

deep discounts for introductory services, as may dance schools and fitness clubs. Before committing ourselves, we normally test-drive a car, and at music stores we can now test-listen CDs. I’ve known lonely economists who shun the personal-ads column, insisting that it is what the economist George Akerlof (1970) terms a “lemons” market. Yet in Akerlof s lemons model we commit to full purchase before learning the slightest thing about the item’s true value. In the companionship market, an hour in a coffee shop allows us to evaluate before “buying”. By presenting the relationship as a series of incremental choices, trusters can gain information to reassess the relationship as it develops.

Another way to alter the structure of interaction so as to lessen the dependence on trust is for the promiser to provide the truster with some advantage to be held until the promise is kept. A tenant, in the role of promiser, gives a security deposit to the landlord. A similar example is the labor contract that defers handsome compensation for high productivity until high productivity has been demonstrated (Hutchens 1989; Carmichael 1989). In the case of products and services, manufacturers and practitioners can offer guarantees and warranties.

An elementary solution is to have the promiser—a building contractor, taxicab driver, or lawyer—provide the service first and bill the truster afterward. This gives the truster some power to withhold payment if she is not satisfied with the experience and throws the burden of initiating litigation into the lap of the promiser. In fact, a functioning court system is itself a means of retaliation against the breaking of promises that are a matter of contract. Indeed, the power to gossip about the promiser is another form of assurance and one that promisers may provide by setting up e-mail chat groups (Hagel and Rayport 1997).

Giving an advantage might re-invent the initial trust problem by creating the possibility of opportunism on the part of the truster. This is a problem, however, that often will be easier to resolve than the initial trust problem. Economist Oliver Williamson (1983, 527) tells of the promiser-king who gives a human hostage to a foreign sovereign. The king gives not his beautiful daughter, whom the foreign sovereign might wish to abduct, but his ugly daughter, because her plainness extinguishes any incentive for opportunism, aside from ransom, in the newly created trust problem. That’s one way a trust problem can be shifted or remade into a more manageable form. In the case of the firm that offers its workers deferred rewards, or the landlord that holds a security deposit, reputational or legal constraints will often give good grounds for trust in the remade situation.

Even the advisor-reaper tension can be significantly resolved by a simple restructuring of interaction. By entering into a prepaid medical plan, the reaper is taken out of the medical relationship. By coupling service delivery with service insurance, prepaid plans, such as those offered by health maintenance organizations, turn the problem of credence characteristics on its head. The logic of prepayment removes the tendency toward excess treatment but also weakens the tendency toward adequate treatment as cost minimization becomes a salient incentive. So the trust problem has been remade, but again perhaps in a more manageable form8. Judging whether inadequate care is given may be easier than judging whether excessive care is given; the victim of inadequate care is more apt to squawk and to require additional attention. And concerns over reputation—both that of the individual practitio-

8 Beito (1997) offers an important examination of the lodge practice experience. He indicates how prepaid plans brought low-cost health care to the poor, bringing with it preventative care, the active patient, the gate-keeper role of primary care and monitoring by the organization. He also explains the unabashed and ignominious suppression of medical commerce by medi-cal societies, who complained that “ [t] he physician is thus placed in the same class as the tradesman.”

5

ner and that of the medical group to which he belongs9 —are likely to give the truster reason for trusting the practitioner in his role as advisor.10

Prepaid medical plans, by the way, were widespread and rapidly growing in the early years of this century, notably among fraternal orders and mutual aid societies, but medical societies, often vested with government power, swiftly put down the “lodge practice”—a term of opprobrium like “scab labor”—and thoroughly suppressed the logic of medical prepayment for another 50 years.11 The potentialities of prepayment are being explored as restrictions on property, consent, and contract are being relaxed.12

Chatting and Getting Face-to-Face CluesIn any sort of personalized service, an important factor is face-to-face contact. By talking we may get a clearer understanding of the content of promises, and public statements of that content can reinforce accountability. In writing about trust, sociologist David Good (1988, 44-45) argues that increased contact makes for “greater certainty as to the intent behind any communication, and, therefore, less likelihood of [ one’s] being misunderstood or exploited.” “[A ]nything which contributes to a greater sense of knowing the other can contribute to the participants’ belief in the possibility of mutual understanding.” In some cases, personal rapport is a part of the experience characteristics, as in counseling, schooling, and childcare. And sometimes, as in medicine, good rapport is necessary to informed treatment.

Talking face to face also helps one to assess character and credence characteristics. As the economist Robert Frank (1988) makes plausible, we get clues about trustworthiness. In my own experience of shopping for a used car I have always paid closer attention to the owner and his home than to the car or any of the bewildering array of components under the hood. If his garage is neat and orderly, with different-sized nails carefully stored in separate jars, and a wide range of gadgets on hand, my faith in the quality of the car grows because he seems like the type to follow proper maintenance. If he looks me in the eye, speaks in a clear voice, and wears a firm mouth, my faith in his character grows because he seems like the type to follow probity.

While 0. Henry remarks that moments of chatting are the raisins in the tasteless dough of existence, I wish to try a different metaphor: they are the ball-bearings in the wheels of trust. Even when trusters have little technical knowledge of the trouble, as for engine repair, they can get an impression as to whether the promiser is trustworthy. For example, it is always a good sign when the promiser takes pains to explain why the trouble is occurring and what the various options for remedy are. An understanding of the trouble usually comes down to a few basic relationships. Explaining the situation helps to inform the truster and creates a measure of

9 Feldstein (1993, 321f) reviews the history of nonmarket barriers to the development of prep aid health plans. These have included denial by medical associations of hospital privileges to participating doctors, and strict regulation of insurance by state departments. State regulations have “required physicians to be a majority of the controlling board of an HMO, required HMOs to permit the participation of any physician in the community, required HMOs to be organized on a nonprofit basis, and prohibited HMOs from advertising their benefits and premiums.” Federal legislation in the 1970s brought federally qualified HMOs under a different set of regulations, including financial requirements, certificate-of-need laws for facility ex-pansion, undesirable reimbursement policies, mandated packages, open enrollment, and premium setting based on community ratings. Several of these regulations have been relaxed (notably in the areas of enrollment eligibility, certificate-of-need, and advertising), but the health-care-organization business remains heavily regulated.10 However, chatting might reveal to a cunning promiser that the truster is naive, ignorant, or powerless.11 Calkins (1928, 23) writes: “A business which advertises, which puts its name before the public and puts its trade-mark on the goods, has entered into a contract with ... its customers.... It is honor bound to justify the advertising. It has given a hostage to fortune.” Ippolito (1986, 15) writes: “A written warranty, for instance, is a more specific articulation of a minimum that the manufacturer is willing to promise the purchaser in the event that problems arise with the product. If the producer does not honor the contract, all consumers who experience the problem will be more certain that a breach has actually occurred. This improves the clarity and diffusion of information about cheating and increases the effectiveness of reputations as a disciplining device in the market.”12 The phrase belongs to the late University of California, Irvine philosopher Gregory Kavka.

6

accountability for the promiser.13 Again, the Internet is making it astonishingly easy to probe the content and integrity of promisers by pointed dialogue.

The austere role of the licensed professional, as well as his high hourly rate, works against trust. Face-to-face trust is probably enhanced where professionals make cut-rate “get acquainted” offers. The building of trust in a promise is usually furthered by a clear articulation of the content of the promise. Hence, a variety of basic informational avenues—advertising, chatting, labeling, guarantees, endorsements, seals of approval—can help traders gain mutual understanding.14 In health care, where many of these media of communication have been regulated, patients are better able to resolve character uncertainty and gain pointed knowledge when the doctor-as-merchant ethic of commerce prevails over the doctor-as-health-official ethic of the professional guild.

Extended Dealings: Time Wounds All HeelsRemaining on Main Street gives rise to the businessman’s reputation—that is, the general opinion of his trustworthiness. Continuance and repetition open up vast institutional possibilities and provide fertile ground for trust. In a sense, our power to damage a promisees reputation or to withdraw from dealings serves as a hostage that we hold against his promises. Career promisers build and protect their reputation, sensing the truth in the saying, Time wounds all heels.15 When not prevented by government, voluntary institutions develop to give bite to the saying, because that arrangement is preferred by all parties except the untrustworthy.

An individual truster that has frequent dealings with a promiser has some grounds for trust, even in the most difficult advisor-reaper relationships. The truster may learn that a treatment was inappropriate, and the malfeasor may lose a customer. Also, in frequent dealings there is face-to-face familiarity, making it diffi- cult for the promiser to keep up deceit. A habit of deceit is a mark of bad character, and bad character has a way of revealing itself no matter how cunning the individual. Deceit is both bad karma and bad business. I am inclined to agree with Montesquieu, Adam Smith, and Friedrich Hayek that commerce elevates manners and probity.16

Alneelish promiser must fear that time will wound him even when dealing infrequently with a truster. Over half of the households in America contain at least two adults, who share their market experience over dinner or during television commercials. Also, trusters share their experience with friends, neighbors and coworkers. By virtue of the Internet we may share information with people we’ve never met.

When someone has a problem and is unsure about how to proceed, usually the first step along the information path is to seek advice.—”Do you think I should see a lawyer for this? How should I choose a lawyer?”—In this manner the experience of another takes on extended life, even if infrequent. When the familiar knows better whether the lawyer is trustworthy, he functions as a knower for the inquirer.

A particular truster has extended dealings with a particular promiser if her dealings are frequent, extended to other trusters, or both. Otherwise she has isolated dealings. People feel most comfortable in extended dealings—13 Sec Smith (1978, 538-39), Hayek (1988, 29-47), HirsChman (1977, 56-66, 107, 128), and McCloskey (1994). A delightful and insightful doux commerce tract that focuses on the role of advertising is Business the Civilizer by Calkins (1928).14 People value predictability. In this respect they resemble dogs: “Seligman, Maier, and Solomon (1971) report that dogs subjected to random and unpredictable shocks of a relatively low intensity suffer much more than those who are receiving higher but predictable charges. Typically, the first group of dogs becomes listless, their physical condition worsens, and their normal interests diminish when compared with the group receiving the greater but predictable punishment” (Good, 32). Schopenhauer (1893 [1851], 162) offers another explanation for the dread of being gypped “Our trust in other people often consists in great measure of pure laziness ... on our own part: I say laziness, because, instead of making inquiries ourselves, and exercising an active care we prefer to trust others ...” We dread being cheated because we dread having to admit our having been lazy.15 Except insofar as people enjoy reading Consumer Reports or going to college (which issues a seal of approval on Commencement Day). And recall what Henry said about chatting.16 Consumer Reports states its position in each issue: “Neither the Ratings nor the reports maybe used in advertising or for any other commercial purpose. Consumers Union will take all steps open to it to prevent commercial use of its materials, its name, or the name of Consumer Reports.” Silber (1983, 31) notes that in the 1950s “[a]ttomeys successfully protected the ratings of the magazine from unauthorized use by commercial interests.”

7

thick and richly diversified extended dealings in trade might in fact merge seamlessly with friendship and community. Trust goes with extended dealings, like good health goes with good nutrition. Both trusters and trustworthy promisers will seek extended dealings, even without being consciously aware of doing so (see Choi and Hilton 1995).

INFORMATION ON QUALITY AND CHARACTER: WHO WILL PROVIDE IT? °Trust and similar values, loyalty or truthtelling, are examples of what the economist would call “externalities.” ... They are not commodities for which trade on the open market is technically possible or even meaningful. —Kenneth Arrow (1974, 23)

As is well known, information on quality has many aspects of a public good... Under such circumstances, inadequate resources will be channeled to providing information. —Hayne Leland (1980, 268)

Trusters dislike being cheated. Furthermore, they dread being gypped.17 They will pay good money for information that protects them.

The provision of quality information is like the provision of parking space. Consider three ways in which parking space is provided in the private sector. First, many motorists provide parking space at their homes by paying for the construction of driveways and garages. Second, independent entrepreneurs provide parking space in lots where motorists park for a fee. Third, shopping plazas, retailers, and office buildings provide parking space where motorists park at no charge.

The private provision of quality information takes the same three forms. Analogous to the private driveway, the truster may gather information by herself, inspecting the product or drawing informally on the knowledge of others. Analogous to parking in a pay lot, the truster may procure information from an independent knower organization, which carries on a trade in generating and conveying information. Analogous to free parking, the truster may acquire information provided freely by the promiser. Because quality information, like parking space, aids not only the recipient of it, but also the party attached to it, either one might, see to its provision. Both trusters and promisers, particularly trustworthy promisers, value information services, and any returns to information providers must come out of gains-from-trade be, twe’en them.18

Do-It-Yourself Information: Informal Channels of Information SharingThe campus housing development where I used to live has a homeowners association that issues a monthly newsletter. In a recent issue there appeared recommendations for a plumber, a painter, an electrician, a Volvo mechanic, a window cleaner, a carpet cleaner, a piano tuner, a woodworker, a brick-layer, a cabinet builder, a nanny, a handyman, a gardener, a furniture transporter, a floorer, two garage door servicemen, and seven house cleaners. My neighbors provided the recommendations, acting individually and giving their own phone numbers for details. People in the neighborhood know each other well enough to doubt that anyone would take a bribe to recommend a lousy nanny or handyman. The recommended individual is quite likely to be an illegal practitioner, even an illegal alien.

The newsletter serves as a sort of community concierge, a file of knowers that efficiently points community members down happy ‘ information paths. The newsletter column exists precisely because there are information

17 “In the absence of special legal protection, the owner (of information] cannot, however, simply sell information in the open market. Any one purchaser can destroy the monopoly, since he can reproduce the information at little or no cost” (Arrow 1962, 151).18 Ippolito & Mathios (1990, 479) examine the effect of the removal in 1984 of a ban on health claims in the cereal market: “The evidence clearly demonstrates that fiber cereal consumption increased once the ban on health-claims advertising was removed. The development of fiber cereals also increased when producers were given the ability to advertise the health features of the products. Moreover, advertising appeared to reduce some of the differences across the population [of cereals], suggesting that advertising may have had its effects by reducing the costs of acquiring information.”

8

problems to be solved. No one would get good-neighbor points or that pleasant helping feeling for helping to solve a non-existent problem.

The newsletter is a kind of local gossip. Gossip arises among family, friends, acquaintances, neighbors, and coworkers. It takes the forms of chatting, group meetings, correspondence, leaflets, bulletin boards, newsletters, local newspapers, and e-mail. Anthropologist Sally Merry writes that “gossip can be viewed as a means of storing and retrieving information.” “It forms dossiers on each member of one’s community: who is a good curer, who can be approached for loans, ... who is a good worker, and who is a thief.” In consequence, “the individual seeks to manage and control the information spread about him or her through gossip” (pp. 275, 279). In the marketplace, promisers do likewise by maintaining quality. A study of consumer complaints against Coca-Cola products found that, of complainers unsatisfied with the company’s response, the median told nine other people about her unsatisfactory experience (TARP 1981, 14).

Fee for InformationWe might view gossiping as a sort of exchange. Exchanging information with acquaintances is one basis for our personal relationships. Information provision by gossip comes to beseen as a trade and as a source of profit. But how can providing information, which resembles a “public good,” be a source of profit?

Information provision can be divided into two stages: generation and conveyance. The generation of information can take the form of testing, inspecting, researching, evaluating, or interpreting. For example, Consumers Union does all of these when generating product ratings in Consumer Reports. Consumers Union makes profits by selling its magazine to trusters. Is its information a public good? Once one person has the ratings, she can indeed share them with her friends and acquaintances—she may even sell her expertise in some manner. But she is prevented by law from reproducing the information and selling it. The information is proprietary and to a good extent excludable.19 If you can protect information at the conveyance stage, then you can appropriate its value at the generation stage.

In the case of Consumer Reports, excludability is achieved in large measure by legal sanctions. Yet often excludability is simply a matter of technical limitations on the part of would-be free-riders. Information conveyance requires information receiving, organization, storing, retrieval, and transmission. For example, credit bureaus like Experian (formerly TRW) sell credit reports to trusters. They make profits by facilitating dealing, just as Manhattan parking entrepreneurs make profits by facilitating shopping. Experian releases valuable information to millions of parties every month, but that does not mean that they can appropriate the value of the information by reselling it to others. Besides proprietary constraints, Experian provides highly individualized information. Its art is in making information complete, speedy, and precise. For someone to free-ride on Experian by entering and competing, she would have to invest in vast data-processing systems. Kenneth Arrow jumps to conclusions when he says that in the absence o special legal protection, entrepreneurs cannot profit by selling in formation.20 Like the private parking garage, the service performed by Experian is largely excludable, although Experian may have to worry about free riding by its two incumbent competitors (TransUnion and Equifax).

In each of the cases of Consumers Union and Experian we see independent knower organizations remunerated by trusters. Consumers Union reports on standardized products, and its conveyance of information is uniform, not individualized. In consequence, it would be damaged by free-riding if its information were not protected bylaw. Experian deals in information of a more sui generic nature, namely credit records, and its conveyance is individual-

19 Beaks, Craswell & Salop (1981, 502f) offer a good discussion of self-disclosure, as well as of imperfections in the quality-information market generally.20 Inside its back cover, in a column called “Selling It,” Consumer Reports exposes dubious and amusing advertising and labeling practices. Blumberg (1989) gives extensive detailed stories about the unsavory side of business; one must question whether his discounting of the role of reputation leads him not only to underestimate the trustworthiness in the marketplace, but also to overestimate the validity of his own source material, since much of it comes from his own students, students who would know Blumberg’s reputation for seeking information that reflects poorly on the marketplace.

9

ized. Its information is protected by the technical limitations of whoever might think of reselling the information, as well as by whatever contractual protections apply and federal law (15 U.S. Code 1681, sec. 607e). Whenever quality information is individualized, the opportunities for reconveying it are limited. Thus some knowers can make money by being hired by trusters to inspect customized security equipment, manufacturing plants, and used automobiles; to give second opinions on medical matters, and to evaluate prospective employees (Rees 1966). And the beloved kibitzer can keep up a steady trade in local gossip because others cannot receive, interpret, organize, retrieve, and transmit information nearly as well as he can.

Kenneth Arrow has pointed out that trade in information is often hobbled by the peculiar condition that the buyer inherently does not know what he is purchasing. Arrow writes, “ [T] he value of information is frequently not known in any meaningful sense to the buyer; if, indeed, he knew enough to measure the value of information, he would know the information itself’ (1963, 946). In the case of Consumer Reports or Experian, however, the buyer does have a good idea of the value of the information she is purchasing, even though she does not know the information itself. Granted, the consumer cannot measure the value perfectly without in fact having the information, but she might know the range of the value, or the expected value. In this respect, Consumer Reports is like The New Republic or a Stephen King novel or a movie ticket.

Information, No Charge: Self-Disclosure by PromisersWhen a knower generates basic quality information on a standardized product of interest to a wide class of trusters, reconveying the information might be easy, and he may go broke trying to sell information to trusters. In that case he goes to work for the promisers (Beales & Salop 1980). Promisers may be keenly interested in hiring knowers—to generate information` about themselves and to convey that information as widely and freely as pos-sible. If a lack of parking spaces would prevent customers from coming to buy, and an independent parking entrepreneur could not meet consumer demand, then the retailer would himself provide space for customer parking, at no charge. Similarly, if a lack of information would prevent trusters from entering into deals, the promiser provides the information. If his quality is high, he has every incentive to self-disclose far and wide.

Pauline Ippolito (1986, 23), a leading student of quality information policy, remarks on how sellers self-disclose:

Low tar and nicotine cigarette sellers have been vigorous in distinguishing themselves from the higher tar brands (going far beyond the mandated disclosures in advertisements). High mileage automobiles often feature this fact in their advertisements. Lower calorie foods (especially in the diet soda and frozen food categories) have been very successful in conveying their superiority to higher calorie counterparts. The same is true for high fiber foods.21

Whenever there is any way of demonstrating or indicating the uses, conveniences, durability, or special pleasures of their products or services, sellers strive to do it.22 They employ salespeople to demonstrate and describe the product, they set up displays, they advertise product characteristics, they recruit the services of referral agencies, they offer guarantees and warranties (which in Grossman’s idealized example (1981) lead to perfect disclosure). Getting the information out to trusters is a necessary input to creating the trust upon which dealings are based. That the provision of information may exhibit “public goods” characteristics is not a curse to promisers but a blessing. They want the information to be freely accessible, just as frontier merchants and landowners want transportation to their town to be freely accessible.

But can the information be trusted? Marketing claims of “New and Improved” often refer to experience or credence characteristics, and the advisor-reaper tension gives cause for suspicion. Promisers often mislead and

21 On the idea that advertising is profitable only for firms that consumers want to continue to patronize, and therefore signals quality, see Nelson (1974), Kihlstrom & Riordan (1984), and Ippolito (1990).22 Viscusi (1978) provides a model of quality certification which yields a happy outcome, in explicit contrast to the unhappy outcome of Akerlof s model (1970).

10

deceive trusters.23 But trusters know their vulnerability and look for solid evidence of trustworthiness. In consequence, promisers who can evidence trustworthiness are rewarded in business traffic.

One sort of evidence is demonstration of traits distinctive to trustworthiness, such as announcing, “Established in 1924,” or engaging in a heavy promotional campaign that would be lucrative only for a worthy promise (Klein and Leffler 1981). Consumers make inferences, if only subconsciously, from the life of the firm, the size of the firm, the extent of advertising, and so on.24 Before the Federal Deposit Insurance Corporation was created to bail out banks, banks had traditionally used large pillars and heavy marble in their architecture to signal permanence.

Another sort of evidence is the word of an independent knower, someone who can evaluate the product but who does not reap gains from its success. If the word is favorable, the promiser will spread it far and wide. Consider the following examples: makers of computers and automotive products mention “editor’s choice” accolades in their advertisements, household products display the Good Housekeeping seal of approval, movie ads reproduce favorable and informative excerpts from the critics, restaurants display favorable dining reviews on their walls, and manufacturers have been known to send retailers reprints of favorable Consumer Reports ratings (Strickling 1965, 18). A transmission shop advertised in my local newspaper that it is “the proud recipient of the Newport Balboa Rotary Club award for honesty & integrity and has been a family owned and operated business since 1960.”

It is good fortune to win an award or be praised by the critics. But sometimes promisers do not wait for fortune to find them. They simply hire a knower to make an evaluation. The knower is remunerated for generating the information, and, if the word is favorable, the promiser invests in conveying it to trusters. For example, electronics manufacturers hire Underwriters’ Laboratories to test and inspect their products and grant a UL mark upon approval. Companies and governments hire Moody’s to rate their securities and use the ratings to market their securities.25 Similarly, promisers assure trusters by advertising in media that police the integrity of the promisers—a strategy first employed against the quackery of patent medicine by Ladies’ Home Journal (Calkins 1928, 49). Another class of knowers paid by promisers, particularly relevant to the issue of occupational licensing, is made up of professional schools, technical schools, institutes, and training programs that grant degrees and certificates. These credentials are then prominently displayed on office walls and listed in curriculum vitae. Transcripts and academic honors give a sort of rating system to the degrees. Each of these organizations grants its own seal of approval.

Research on seals of approval has suggested that seals like the UL mark are not really understood by consumers and do not significantly enhance the consumers’ confidence in the product (Parkinson 1975; Beltramini and Stafford 1993).But the research fails to appreciate the importance of pointed knowledge. Researchers show test subjects advertisements with and without seals of approval, and see if subjects have greater confidence in the ads with seals of approval. Unlike genuine prospective buyers, however, test subjects do not have an interest in the particular products advertised, and hence do not have the incentive to gain pointed knowledge about relevant signals of quality. The research also says that consumers poorly understand seals of approval, because they do not know on what basis the seal is awarded. But again, pointed knowledge for the consumer is knowledge of whether products with the seals tend to be satisfactory, not formal knowledge of how seals are awarded. Knowledge of how seals are awarded might be known to only a few, but those few may provide the base upon which an inverted pyramid of divided knowledge is sustained,

23 “[T]he ban substantially increased the cost to firms of introducing new low-tar brands and the cost to consumers of obtaining information about these newer brands, thus slowing down the movement to these lower-tar cigarettes” (Schneider et al. 1981, 610).24 BBB 1992 Annual report, p. 2.25 Modern Healthcare, August 31, 1992, 40.

11

FIGURE 1 CLASSIFICATION OF KNOWER SERVICES

Knower is remunerated byTRUSTERS PROMISERS

Know

er e

ngag

es in

info

rmati

on GENERATION Hired inspectors(for buildings, automobiles)Letters of recommendation DoctorsFinancial advisors Hired investigators American Automobile Association

Credential givers (universities, institutes, training programs)Underwriters’ Laboratories American Dental Association Good Housekeeping Security ratings (Moody’s,Standard & Poor’s) Securities underwriters Financial and accounting auditsNotary publicLetters of recommendation Orthodox Union (kosher foods)Internet seals of approval (TrustE, Cyber Patrol, Safesurf, Verisign,BBB Online)

GENERATION ANDCONVEYANCE

Consumer Reports Dun & Bradstreet Industry newsletters Hobby, product, and news publications Restaurant and moviereviewsEmployment agencies BrokersInternet chat groups (eBay)

Better Business Bureau Medical data banks Employment agencies Brokers (securities, realestate, produce, art,collectables)

CONVEYANCE Gossip, e-mail Consumer credit bureaus Referral services Advertising firmsSigns, labels, packaging, displays, sales help Web pages

making the seal an effective signal to those farther up the pyramid who use only very limited pointed knowledge. Furthermore, advertising credibility may not he the relevant test of a seal’s value: the seal may be most important to the retailers and other middlemen who decide whether to carry the product.

A Classification of Independent Knower OrganizationsTwo distinctions aid us in thinking about knower organizations: first, whether the knower is engaged in information generation or conveyance (or both); and second, whether the knower is remunerated by trusters or by promisers. Using the two distinctions we get a classification scheme as shown in Figure 1.

The left-hand column of Figure 1 shows knower organizations that are remunerated by trusters. Those that generate highly individualized information, like hired inspectors and people who give second opinions, do not face free-rider problems-simply because the information buyer cannot hope that someone else will pay to have the pertinent information generated. Knowers that generate information about standardized products—information of interest to many trusters—generally make money from conveying the information. To do so they must enjoy a measure of exclusion. Their information may be protected by law (e.g., Consumer Reports ratings, any type of copyrighted review), or difficult to re-convey (e.g., credit reports, gossip). The right-hand column shows knower organizations that are remunerated by promisers. Promisers may pay them for generating a seal of approval (certifiers, inspectors, credential givers), for conveying information (referral agencies, advertising firms), or both (information bureaus).

“Prologue,” by Consumer Health Services (CHS)In nine metropolitan areas, covering 16 percent of the U.S. population, one can dial 800-DOCTORS and shop for medical attention. “Prologue” is an information service provided by Consumer Health Services, a for-profit corporation. CHS has 325 full-time employees, 200 of which are Prologue telephone “counselors.”

12

The service is a fine illustration of pointed knowledge. Practitioners fill out a 20-page questionnaire that includes 1300 information variables, or “searchable codes,” including specialties, treatments performed, equipment, degrees, certifications, practice philosophy, sex, location, hours, and fee ranges. The caller tells the counselor of her specific needs, and the counselor puts the pointed requests to the computer using the searchable codes. With the counselor acting as intermediary, the caller interacts with the computer to produce a menu of practitioners. Once the caller selects a practitioner, the counselor places a conference call to set up an appointment and asks for the fee range of the services sought. CHS gets a fee from the doctor for an initial appointment, or from a sponsoring hospital. Prologue takes 100,000 calls per month, and arranges 25,000 appointments per month. About 8,500 physicians participate in Prologue, 1.3 percent of the country’s physicians.

Prologue is mainly a match-making service, not a reputational mechanism. A refined match-making service improves mutual understanding of expected service and prices. The acquisition of pointed knowledge is improved, and the credence element of the relationship is reduced.

There is a reputational component to Prologue. First, the organization verifies some credentials. More significantly, it asks users to fill out response cards that ask about the doctor’s communication style, the amount of time spent with the patient, the fees, treatment by the staff, and so on. According to a CHS representative, CHS terminates a practitioner’s participation any time he “receives serious negative consumer feedback ... or fails to solve problems which have produced an on-going record of negative” feedback. Thus, being included in the Prologue database is, to some extent, a seal of approval.

Looking through The Consumer Sourcebook (Wilson 1989), I find scores of organizations that engage in at least one of the following activities: testing and rating products, certifying products, giving credentials, and operating reporting bureaus. A great many help to resolve disputes, provide referral services, and provide consumer-information literature. We learn of such institutions only in the process of solving a particular problem.

Ideas for Further Research on Knower OrganizationsWe can draw several distinctions between independent knower organizations: (1) whether they are remunerated by trusters or promisers; (2) whether they generate or convey quality information; (3) whether information is excludable (or proprietary); (4) whether they report on standardized or individualized services; and (5) whether information is based on past dealings or independent evaluation.

Numerous questions worthy of exploration arise: What sorts of dealings have depended historically on the existence of independent knower organizations? How much is a seal of approval worth to a promiser? Do trusters come to recognize credentials and seals of approval? Is competition among seals of approval desirable, or does it create bewildering cacophony? Do trumpeted seals of approval always signify satisfactory quality? Do seals of approval act as barriers to entry? Do knower organizations foster collusion? What leads members of BBII and similar organizations to join and cooperate? How do organizations that develop standards interface with the legal system? Do these orginizations remain impartial? What keeps them honest? Where information is based on past dealings, what damage is done to privacy? (See Larson 1992 for a critical view on privacy.) What can we learn from knower organizations about undercover tactics as a way to monitor and control behavior? What types of knower organizations tend to act as rent-seekers for their promiser-supporters (for example, the A.M.A.)? What types of organizations tend to act as rent-seekers, or mere political crusaders, for their . truster-sup- porters (for example, consumer-interest lobbyists)? Will encryption and the Internet enable everyone to pursue detailed and interactive information paths about product safety while sitting at their personal computers?

Ways of Apprehending UntrustworthinessTrustworthy promisers have every incentive to self-disclose, and they employ a wide variety of means to do so. What about the untrustworthy? They do not strive to self-disclose; in fact, they have a special incentive to deceive. How do trusters apprehend the untrustworthy?

13

One way is to ask for a public elucidation of the content of the promise. If they fudge and prevaricate, suspicions arise. Perhaps even more telling are the accolades, coveted seals of approval, and glowing endorsements that are not. When we view a curriculum vitae, a meagerness of distinctions will make itself evident and lead us to doubt outstanding talent. Similarly, trusters remain wary when they do not hear any of the wide variety of horns that trustworthy promisers blow in self-disclosing. It is precisely because the horns are denied to untrustworthy promisers that they are effective signals of quality.

Another way to apprehend untrustworthiness is, of course, to engage knower services remunerated by trusters. Hired inspectors, Consumer Reports, Dun & Bradstreet, Experian, Siskel & Ebert, and the neighborhood gossip all report on the trustworthy and untrustworthy alike. Research finds that customers’ negative word-of-mouth is more influential than positive word-of-mouth (Arndt 1967; TARP 1981).

A third way of apprehending untrustworthiness is forged by competitors.

Competitive ExposéMost of the knower organizations listed in the right-hand column of Figure 2 are institutions that promisers use to self-disclose. An exception is the Better Business Bureau, which is mainly aimed at exposing unsatisfactory business practices. Trustworthy businesses value the BBB not for the praise it gives to them—problem-free files are merely reported as “satisfactory”—but for the criticism it makes of their untrustworthy competitors.

Promisers expose the poor characteristics of competitors’ products, if only by insinuation, in advertisements, sales demonstrations, and marketing literature. Before the imposition in 1971 of restrictions on cigarette advertising, advertisements for low-tar brands sometimes pictured rival brands and listed the tar content beside each. Many researchers think that the restrictions have inhibited the market for low-tar cigarettes.26 Competitive advertising is a great service to trusters, as it helps them discover product differences and the validity of product claims.

Besides exposing poor quality in products, competitors expose misleading marketing practices. In 1984, when leaded gasoline was disappearing gradually from the market, I noted the following radio advertisement for Getty gasoline:

Has this ever happened to you? You’re cruising along on a near-empty tank when you see a gas station ahead with a big sign featuring an invitingly low price per gallon. So you pull up to the pump and that’s when you learn that the price on the sign was for leaded gasoline. The unleaded your car needs costs a lot more. But, what the heck, you’re at the pump, so you fill up. And pay up...

You can avoid being misled, the speaker continues, by looking for Getty, since Getty is now offering unleaded gasoline only, and at a price below most other brands. Getty is selling itself by reminding the listener of her unhappy experience with the competition.

Advertising executive E.E. Calkins (1928) wrote of an early case of competitive expose in advertising:

Dr Lyon’s Tooth Powder and Colgate’s Dental Cream are both using their advertising space to offset undue claims instead of stretching them further. That is one of the values of advertising. It will correct itself. The lying advertisements will find themselves surrounded by truth and will be forced back in line by the weight of public opinion. (Calkins 1928, 284)

26 Undercover work by government would seem to be a powerful tactic of quality control, but is rarely suggested for health care and so on. It would seem that private organizations could sell ratings based on undercover monitoring to auto-mechanics or doctors. Why don’t such organizations exist in these fields? The possibilities are: the difficulty of establishing credibility on the part of the rating organization (Beales & Salop 1980, 240), the high-risk, front-loaded investment in winning recognition for the system, and regulatory barriers or crowding out by government quality-control programs.

14

Of the 65 advertising challenges resolved by the BBB’s National Advertising Division in 1992, almost all of which dealt with the truth and accuracy of advertising claims, 72 per cent were brought by competitors.27

The federal government has taken advantage of competitive exposé. The Department of Health and Human Services has opened a toll-free number for anonymous reporting of Medicare and Medicaid fraud.28 In a similar vein, the health-care economist Paul Feldstein (p. 308) remarks that health care organizations are “likely to carefully monitor their competitor’s claims and report any misinformation to the FTC.” Government services that supplement private activity, by taking advantage of incentives for self-disclosure or competitive expose, or by engaging in undercover monitoring,29 strike me as much more promising ways to mitigate trust problems than dirigiste restrictions on property, consent, and contract. If the problem is that we as consumers are uninformed, why don’t governments try to inform us, rather than eliminate some of our freedoms?

In the areas of health care and pharmaceuticals, competitive expose is restricted—drug manufacturers are not permitted to report findings about their own products, much less their rivals’. A robust and lively arena of self-disclosure and competitive expose, which some regard as freedom of speech, would help trusters to convert general knowledge into the opportune, pointed knowledge they need to guide their behavior (Ippolito & Mathios 1990, 479; Russo et al. 1986)). This function of advertising is noted by the consumer advocate Mark Green (1980, 279), who, in criticizing restrictions on lawyer advertising, says that other liberalizations in the lawyer market “are like one hand clapping without advertising.” In his famous paper on the market for lemons, George Akerlof (1970, 495) suggested that sometimes “dishonest dealings tend to drive honest dealings out of the market.” Freedom to engage in self-disclosure and competitive exposé is one of the freedoms that make just the reverse true.

We have reviewed numerous ways in which market participants generate information to answer consumer wariness and remedy trust problems. Now I turn to a consideration of the forces that tend to shape firms and markets so that, by their very structure, transactions are embedded in extended dealings.

MARKET FORMS AND EXTENDED DEALINGSThe idea that ‘brand name’ and reputation will perform [the quality control] function is just so much fluff ... —Robert Evans (1980, 250).

The Free-rider Properties of Extended DealingsGossip, letters of recommendation, newsletters, data banks, consumer survey literature, information reporting bureaus, and referral agencies all make for extended dealings. These institutions serve not only those trusters who make use of them, but also those who do not. Although a promiser often knows whether a particular truster has frequent dealings, he rarely knows whether the truster has extended dealings. The promiser does not see the truster report on her experience, and therefore, except in rare cases like the hapless motorist who pulls off the interstate for sudden repairs, the promiser must treat every truster as one that might be extending information to others. Thus many lackadaisical trusters gain by the presence of persnickety trusters.3° Dealing extension indeed exhibits positive externalities among the set of trusters, and there is an argument for having government facilitate or perform such services (Beales and Salop 1980). But even though this sort of free-riding among the trusters does occur, the informed portion of the clientele does create a margin of punishment and reward, a margin that favors the trustworthy in the contest for commercial prosperity.

27 Except when they are being persnickety while the lackadaisicals are waiting in line.28 Economist Eric Bond (1982) studied the market for used pick-up trucks, looking for lemons-market results, and found none. He reports (p. 839): Pick-up “trucks that were purchased used required no more maintenance than trucks of similar age and lifetime mileage that had not been traded.” It would be interesting to learn whether used vehicles purchased from dealers require less maintenance than those purchased from isolated individuals.29 Feldstein (1993, 327) writes: “One interesting development in this regard is the purchase of un iversity teaching hospitals by investor-owned chains; such investments are intended to provide a quality name recognition for the firm’s health care activities.

15

Notice that extended dealings might benefit even those trusters who, like the evident out-of-towners, are known by promisers to have only isolated dealings. In the marketing of a standardized product, trusters cannot be dealt with selectively. And even for services, it may be that protocol and the force of habit—including moral habits—keep the promiser honorable even when he knows that cheating would go unpunished..

Trustworthy Promisers Cultivate Extended DealingsInformation sharing is a precautionary tactic used by wary trusters, but it is not necessarily unwelcome by promisers. It is unwelcome only by untrustworthy promisers. Trustworthy promisers welcome information sharing and, where permitted by law, tend to organize themselves to facilitate and expand the extension of dealings.

There are two ways in which a trustworthy promiser gains reputation by having a large base of trusters with extended dealings. First, when extended trusters are pleased by the service, they increase their own patronage and spread the good word to other potential trusters. Dealing extension is a sort of publicity that can attract new, discriminating customers. In his book Industry and Trade, Alfred Marshall (1927, 297) referred to “that highest form of advertisement, which comes from the recommendations of one customer to another; and from the inducements which dealings with one department offer to dealings with another.”

Second, promisers who enjoy a large extended base attract new, nondiscriminating trusters merely by that fact. Less persnickety trusters are attracted to a promiser with a large extended base because they know that such a promiser has strong reputational incentives to make good, and that trustworthiness has probably been a means by which he achieved a large extended base. They may save themselves the trouble of taking steps down an information path by simply trusting the promiser with a familiar name and a large clientele.

Since trustworthy promisers sec that extended dealings bring them good offices, they try to arrange the business to have frequent and extended dealings with their trusters. By bringing an array of services under the umbrella of his name, a trustworthy promiser encourages a pool of fresh trusters to engage his services and the opportunity for satisfied trusters to trust him again.

The Umbrella of the Brand NameIn the late 19th century, as transportation systems and mass production created a national market in America, consumers confronted, according to consumer historian Norman Silber (1983, 3), “a profusion of unstandardized packaged goods ... [and] unfamiliar selling and processing techniques,” making it hard for them to judge such qualities as “the freshness of food or the durability of clothing.” Silber tells how the market responded:

To ease the minds of customers about problems of quality, reliability, and safety, manufacturers and advertisers appealed to consumers to buy according to brand names. National Biscuit, Heinz Soup, Armour Meat, Standard Oil, and other companies placed one banner on many different products. The consumer who found one product of a brand to be satisfactory, those companies suggested, could assume that all other products also would be suitable. (Silber 1983, 3).

16

A brand name is a way of gathering together an array of services that make for frequent dealings. The array will be shaped by finding a fit with the tastes of the clientele (as well as by scope economies). Game theorists tell us that repetition makes for good offices; hence promiser§ try to enhance repetition.

A machine-tool company such as Black & Decker makes hundreds of different products, but its customers will generalize to some extent about all of them based on their experience with only a few (see Goodman, Broetzmann, and Ward 1993 on customer satisfaction and brand loyalty). By enlarging its product base the company creates frequent dealings with many of its customers, giving them a better opportunity to evaluate its trustworthiness. In this way, Black & Decker becomes an institution-.providing the essential service of trust, as well as tools. The inventor-genius may create, de novO, in his basement workshop a fantastic new tool, but he has not produced a great product until he has created trust. To achieve the latter he must collaborate with those who are practiced in striving for and attaining trust; he may find it to his best advantage to sell his invention to Black & Decker and let the firm offer it under the umbrella of its brand name. In a sense, Black & Decker is the expert knower that tells the truster that the inventer’s new gizmo is trustworthy. Black & Decker is not only a manufac-turer, distributor, and advertiser, it is also a knower that grants its own seal of approval. A parallel example is a pharmaceutical company that evaluates new drugs and grants its seal of approval.

Dealers Make for Extended DealingsBesides generating extended dealings with consumers, Black & Decker is at the center of a starlike pattern of dealings with scattered inventors. Consider the similar case of the used car dealer. The used car dealer might have only isolated dealings with the sellers of used cars (like Black & Decker has with some inventors). The car dealer knows all about cars, so, unlike ordinary individuals, he deals on an equal information footing with the seller. By gathering up a stock of used cars from an array of isolated sellers, the dealer produces a fixed lot of cars and a basis for extended dealings with. buyers. Although he does not have many frequent customers, the dealer is a fixture in the community and word of his business practices gets around. A buyer gets to know her car intimately, and if the experience characteristics disappoint her, she will have ample opportunity to spread the news. Also, the dealer can offer guarantees and warranties. The dealer, then, besides reducing transaction costs and upgrading the commodity, transforms a series of isolated dealings—dyadic matching between many sellers and buyers—into a series of extended dealings: a fixed seller dealing with many buyers (Kress 1990 Figure 2 shows the starlike

17

pattern of interaction that the dealer has with sellers. In this manner he creates trust, and he receives a payment for doing so.31

In an article entitled “Middleman as Experts,” economist Gary Biglaiser (1993) models how middlemen can ease information problems if their sunk investment in becoming an expert can be demonstrated. He gives the following example: “Jewelers display that they are graduate gemologist, a certification process that takes at least six months and costs at least $2,000” (221). The infrequent buyer feels that she can trust the certified expert because his cre-dential gives him credibility as a knower and is a costly, irreversible investment. Those two conditions give the middleman both manifest competence and manifest earning-power in the business, provided he does not tarnish his reputation. Hence, it is common knowledge that he has an incentive to protect his reputation, because he will be around to reap the returns of honest dealings (Marshall 1927, 270; Klein & Leffler 1981; De Long 1991; Nichols 1998). Biglaiser gives other examples: “Coin and stamp dealers display to the public that they belong to dues-paying professional societies and are certified numismatists and philatelists ... Many used-car dealers train mechanics to check and maintain car quality.

If the dealers cheat customers and go out of business, then the investment in their employees’ human capital is lost.”

Reputational Nexus and the MiddlemanWhen my neighbor recommends to me his familiar, for example, a gardener, I have greater trust in the gardener not only because my neighbor acts as a knower of the gardener’s trustworthiness, but also because my neighbor’s frequent dealings with him enhances his trustworthiness in his dealings with me. That my neighbor has relationships with both the gardener and with me embeds my relationship with the gardener within a reputational nexus, in this case a triangle with extended dealings as each of the three sides. A reputational nexus is a constellation of extended dealings. Reputational nexuses exist in every area of life •—the family, the church, the social club, the neighborhood, the workplace, and the marketplace. Our social linkages, including all our chatting, create a variety of nexuses that hang together to form one vast netting. Social network theorists figure that any pair of adult Americans can he linked by three or fewer intermediary acquaintances (Pool & Kochen 1989, 16).

Trust depends on both the thickness and the pattern of the links. In the marketplace, the providing of reputational links, as a basis for trust, is a service that businessmen can specialize in and make a living from. The livelihood of the middleman often depends on providing and thickening reputational links. One function of the middleman is to create a bridge of trust between two traders. As the economist Janet Landa (1994, 125) puts it, “the middleman ... mediates between traders . . . who do not trust each other but mutually trust the middleman.”

When a motorist pulls off the interstate and into Joe’s Garage for sudden repairs, she will have isolated dealings with Joe and be highly vulnerable to the advisor-reaper tension. The motorist would do better to pull into Midas, Shell, or Mobil, because, if the local Midas franchisee cheats her, it faces the prospect of punishment—not from her (the motorist), but from the franchisor. Franchisors police the service and probity of their franchisees using “mystery shoppers,” audits, inspections, and complaint investigation because they do have to fear that customers will harm them by injuring the reputation of the franchise brand name. Even if the motorist doesn’t reflect on the advantages of Midas over Joe, when faced with the choice she is instinctively more wary of Joe.

18

The franchisor provides a reputational bridge between the motorist and the franchisee. When the serviceman wearing the Midas shirt and cap approaches us, he is not our connection to Midas. Midas is our connection to him. Midas is like a friend, and the serviceman is the friend of a friend. Although it is mutually understood that the motorist and the Midas franchisee will be interacting only once, the motorist has extended dealings (of an indirect sort) with the franchisor, who in turn has extended dealings with the franchisee. Thus the franchisor creates a bridge of trust between a truster and a promiser that meet only in passing (see Figure 3). A franchise operation succeeds partly by capitalizing on product familiarity and low-cost replication of a successful formula, but also by developing a reputational nexus that permits it to capitalize on trust. Franchised motels, restaurants, and convenience stores also capitalize on trust in infrequent dealings. Other chains, like the Sylvan Learning Center or Fred Astaire Dance Studios, offer a non-standardized, frequently used service, but nonetheless gain trust by patrons and knowers recommending the service to friends living in other areas. One wonders whether, in a free and open field, we might see health, legal, educational, and child-care services provided by firms using brand-name, franchise-like arrangements.32 In a sense, we already have extensive franchising in the field of prayer and spiritual service.

The firm, the chain-store, and the franchise are all different species of contract nexus (Rubin 1978). Simple contracting can achieve trust in much the same way that a chain-store does. The health care organization contracts with physicians and hospitals:

the patient has extended dealings with the HMO, and the HMO has extended dealings with the physician. The reputational-bridging role of the HMO ranges over a contractual continuum, from employment within the firm (the staff model), through intermediate stages (group practice, individual practice associations), to selective contracting with health care providers (Wagner 1989).

The role of the reputational bridge is best demonstrated by the retailer who, a hundred times daily, serves as the link between a consumer and a producer. Many of the matches between consumers and producers are irregular—as in the case of a consumer purchasing a washing machine or an ulcer medication—but the consumer has extended dealings with the appliance store or pharmacy, which in turn has extended dealings with the producer. By carrying the appropriate array of products, middlemen can make a profit by creating trust for the consumer with an irregular need of a specialized commodity. The middleman can enhance his reputation for marketing high-quality products by devising ways of extending his dealings with each side of the link. He will carry numerous products produced by a reliable manufacturer and offer sets of commodities that satisfy an inter-related cluster of customer interests, making for extended dealings.

The Middleman Also Acts as KnowerBesides straddling two distinct extended relationships, the middleman also acts as a knower. The retailer specializes in knowing good products from bad and carries out this function in any of the following ways: by recognizing brand names and seals of approval, by studying the information provided on labels and packaging, by conducting his own tests and inspections, by keeping track of customer complaints and purchase-returns, by hiring testing services, by following trade literature or consumer literature (Strickling 1965, 34, 42), by finding out whether other retailers carry the product, and by chatting with colleagues at industry meetings. Furthermore, the retailer judges simply by the pattern of sale. If a product moves from the shelves only when advertising or coupons push it, the retailer is led to discontinue the product, partly because it is not generating revenues, but also because he reckons that it does not satisfy even those who buy it. This process benefits consumers who would have tried the product and been disappointed.

In his role as knower, the middleman works in an information that is often too costly for the consumer herself to gather and judge (Pashigian and Bowen 1994). The wary consumer finds it too troublesome to traverse the information paths herself. In a sense, the premium she pays to the middleman, whether he is an established

19

retailer, a brand-name manufacturer, or a contracting organization like an HMO, is a fee for the luxury of being both uninformed and covered.

The information service provided by retailers is welcomed also by manufacturers of good products. A good commodity has no entitlement to a strong market; it must concede its dependence on the services and institutions that permit its goodness to be recognized. By creating trust, effective middlemen, such as Nordstrom’s clothing stores, sell more of the product and at higher prices. Manufacturers respond by seeking to have their products carried by such middlemen. (Notice that, like Landa (1994), I am using the term “middleman” in a sense that is much broader than the common usage, of one in between the manufactuer and the retailer.)

The two end points of the reputational link, the manufacturer and the consumer, both have an incentive to avoid the trustless prisoner’s dilemma outcome. Among the diverse, complex, and imperfect institutional experiments that take place in a free and open field, trusters and promisers will favor and sustain those experiments that overcome the trust problem. And the knowers and middlemen who compete for their favor will also seek to pro-mote trust, because ultimately their cut comes from happy trusting outcomes.

CONCLUDING REMARKSlt is no accident that social life is arranged so as to minimize the occurrence of one-shot prisoner’s dilemmas. —Brian Barry and Russell Hardin (1982, 385)

I have tried to develop two claims. First, the avenues to trustworthy dealings are multitudinous and open-ended. Should you trust Mr Dealer? To help yourself decide, you can get to know him, ask him to elucidate publicly the content of his promise, look into his credentials and seals of approval, try out his service on a trial basis, ask for a warranty, ask a neighbor about him, get a second opinion, consult an information bureau or a rating organization, find out what his competitors are saying about him, or have someone that you trust arrange your dealings with him. The actual institutional forms that such aids can take are far more numerous than we can ever hope to know. If, after all your efforts, you cannot come to trust Mr Dealer, then don’t! Look for someone else who gives better grounds for trust.

The second claim I have tried to develop is that voluntary institutions—including knower organizations, firms, market forms, and social networks—evolve not only io provide quality and safety, but to provide quality and safety assurance. These institutions sometimes avoid the need for trust by equalizing information or by making service a matter of explicit contract. Sometimes they shift the trust problem to new ground, but ground where the problem is more manageable. A failure to provide quality and safety assurance opens up opportunities for individuals and firms to profit by providing what is lacking. Bad outcomes breed their Avn remedies.

In 1963 Kenneth Arrow wrote: “It is the general social consenus, clearly, that the laissez-faire solution for medicine is intolerable” (p. 967). Yet the only rascal in the laissez-faire story is the intrustworthy promiser. A vote for coercive government measures must rest on the belief that this rascal cannot be adequately oiled by a sensible tort system and the open-ended, voluntary Tocesses generated by resourceful middlemen, qualified knowers, trustworthy promisers, and wary consumers who purposively seek ointed knowledge.

20