I. THE PROBLEM

Strong Privacy and Public Key Encryption

The increasing popularity of the Internet has reawakened an old concern – privacy. During the days of traditional mail i.e., letters in envelopes, there was a concern about keeping the contents of letters private. When the telephone was invented, there was a concern about people listening to the conversations of others. The Supreme Court of the United States has recognized privacy as a right derived from several portions of the Constitution[1], signaling the importance of privacy in our society. The right of privacy has been used to protect and define a number of other rights such as the right to marry, the right to use contraceptives, and the right to have an abortion.

Privacy has long been something our society values. We have spent much time trying to protect our privacy. We have developed a complex system of laws designed specifically to preserve our privacy in places such as our homes and on our persons. Search warrants are required (for the most part) before the police can search someone’s home; wiretap orders are required before the FBI can listen to someone’s phone conversation; rape shield laws protect the victims of sexual assault from disclosure of their previous sexual behavior in court under most circumstances. But protection of our privacy is far from complete when it comes to the Internet. The obvious lack of control over information on the web has caused some great concern among consumer rights advocates and the government.

I’ll begin with an example. Suppose that in order to buy a book from Bookstore.com, a consumer has to supply Bookstore with his name, address, credit card information, email address and phone number. Bookstore claims that it needs all of this information in order to process the order, collect payment, and deliver the goods. Granted, there currently is no better way to do this. Bookstore cannot deliver to its customers without knowing where to reach them.

But why does Bookstore also keep track of what each of its customers buys, how long he has been browsing through its site, and what types of things he is looking at? Bookstore claims that this is all useful information that it keeps in order to improve the customer’s shopping experience. There is, however, another use for this information that Bookstore is not likely to disclose – the information is sold to the highest bidder.

Suppose Fishmart sells fishing equipment. Fishmart is interested in increasing its customer base. It contacts Bookstore and asks to buy the names, addresses and email addresses of all of Bookstore’s customers who have purchased books about fishing within the last sixty days. Bookstore needs to reach profitability as soon as possible, so it sells Fishmart this information. Fishmart can now mail those customers its catalogs and email them advertisements in hope of increasing its own sales. Bookstore has an incentive to maximize the value of its database so it collects as much data as possible about each of its customers in order to serve a larger population of businesses willing to pay for the information. Business may be good for Bookstore, but the customer is now burdened with loads of junk mail and spam.

I will admit that the above example is not entirely a figment of my imagination. The bookstore that I describe above is Amazon.com. Amazon changed its privacy policy in 2000 to include information about its customers as an asset of the business.[2] The concerns of such violations are real, and the fact that companies are willing to do sacrifice the privacy of the customers in order to make some money is disturbing.

So how do individuals protect their privacy in the absence of any legislation? The solution is strong privacy. Strong privacy is a term that describes how privacy can be protected with the help of technology. Strong privacy allows people to remain completely anonymous online while at the same time allows them to prove that they are who they claim to be. The technology behind strong privacy is public key encryption.

Public key encryption is conceptually simple. Each person has a pair of keys – public and private. These keys are a long string of binary digits: zeros and ones. The keys are generated by commercially available encryption software. The individual keeps his private key absolutely secret. No one else in the world can know the private key. The public key is widely published. Either key can decrypt a message encrypted by the other, but the same key cannot decrypt a message encrypted with it. The system works as follows. Anyone seeking to reach me writes an email and encrypts it with my public key. I, being the only person in the world in possession of my private key, am the only one who can decrypt the message. If I want to write someone else, I encrypt my message with that person’s private key. The result is that anyone who intercepts either of these messages cannot decrypt them because he lacks the respective private keys necessary for decryption.

The Benefits and Costs of Public Key Encryption

Public key encryption carries with it a number of benefits, but it also creates a number of serious problems. The first and arguably most important benefit is that anonymity is achieved through which absolute privacy is obtained. Consumers can go online and conduct their business without worrying that someone is monitoring their communications. Messages can be sent knowing that only the intended recipient will be able to read them. Sensitive information can be sent securely and with complete confidence that it will remain secret until it reaches its destination.

Anonymity, however, eliminates the ability to know with whom you are dealing. How do I know that the message I am sending is really going to the person I want to send it to? The solution to this problem is known as a digital signature. Since I know that the owner of the private key matching the public key that I intend to use is the only person able to decrypt my message, I can send a test message to verify his identity. If the response comes back positively, I can be confident that the message I send is going to the correct person. The ability to decrypt a message encrypted with his public key is a signal to me that this person is who he claims to be. Since the keys are unique, only one person can be connected with each pair, which allows me to verify that the other person is really the holder of the private key.

Several states have enacted laws that recognize digital signatures as a valid substitute for traditional signatures that are required for certain types of transactions. Among them is California. California’s law is perhaps the most widely accepted in the nation, probably because it is written broadly. California’s Government Code § 16.5(d) defines a digital signature as “an electronic identifier, created by computer, intended by the party using it to have the same force and effect as the use of a manual signature.” Section 16.5(a) of the Code states that a digital signature has the same force and effect as a manual signature if it meets a number of criteria including: uniqueness to the person using it, it is capable of verification, and is controlled only by the person owning it. California clearly adopts the validity of digital signatures.

The Uniform Electronic Transactions Act of 1999 also recognizes the validity of use of digital signatures in commerce. Section 7(d) of the Act states that “[I]f a law requires a signature, an electronic signature satisfies the law.” An electronic signature, as defined in section 1(8) includes a digital signature. Section 13 of the Act forbids the exclusion of a digital signature from evidence solely because it is in electronic form. The Act has been adopted by 39 states, including California, as of the end of 2001 according to the Uniform Law Commissioners web site. There is clearly a trend towards adoption of digital signatures as a valid means of verifying identity.

Another problem created by public key encryption is the possibility of the creation of illegal enterprises. Suppose the mob decides to go into the business of professional assassinations. The mob opens a site online and publishes its public key in major newspapers and online. For the sake of argument, I assume the mob is able to keep its identity secret when it registers its domain name. The mob’s ads say the following: “Is there someone you want to off? Send us a message at [public key] and we will take care of it. Satisfaction guaranteed.” There is no way for anyone to know that the mob is behind this business because the identity of the mob is secret. There is also no way of knowing who sent the message because his identity is also secret. The result of anonymity is to allow professional criminal organizations to operate online, and customers to patron them, without any effective way of stopping them.

Enforcement is not only a problem in the extreme case discussed above. Public key encryption allows potential terrorists to organize without jeopardizing their identity or their whereabouts. There are those who argue that a little intrusion into one’s privacy is a small price to pay for national security. The FBI’s carnivore system currently scans email messages in search of key words that could tip the FBI as to certain kinds of threats. This system will become obsolete if the messages are encrypted. The ease with which terrorists can form and execute their plans increases greatly. Also, the likelihood for mass fraud increases if con-artists are able to hide their identities, or give themselves any identity they want, and take advantage of the relatively low cost of email to spread their lies. And as ecash becomes available, it will be even easier to collect payment from such scams.

Finally, and somewhat most worrisome, is that our ability to apply the current system of law to online transactions becomes impossible. The scope of this paper will be the impact of strong encryption on the enforcement of our current contract law. But before I focus on the enforcement of contracts, I believe that a quick overview of the broader system of law, as it currently exists, is appropriate.

The Current Legal System

How does our current system of law award damages to tort victims or punish criminals? To oversimplify the process for the sake of clarity, in a tort action, a plaintiff files a complaint in a court with proper jurisdiction. In the criminal system, the state is the party that files the complaint. In either case, the defendant is then served with notice of such filing. The parties go to court and a judge, and sometimes a jury, decides the outcome of the litigation.

The court will examine all of the relevant evidence before making a final determination of guilt or innocence. The current evidence code relies heavily on physical evidence such as a written document for a breach of contract action, or a weapon for a murder charge. Courts will sometimes examine the intent of the parties or the level of the defendant’s premeditation. Because of the inherent difficulties in determining someone’s state of mind on a previous occasion, the system relies on certain presumptions that keep the process from stalling. Some of these presumptions include: (1) “A letter correctly addressed and properly mailed is presumed to have been received in the ordinary course of mail;”[3] (2) “Money delivered by one to another is presumed to have been due to the latter.”[4]

The appropriate remedies in the tort and criminal systems are different. In a tort action, a party that is guilty will usually pay monetary damages. Sometimes, when equity so requires, the court will award an injunction in favor of the winning party. In a criminal context, the proper remedy is a jail sentence, or in the rare instance, death.

The current law assumes that the parties entering into agreements know each other’s identity. This assumption is still correct for the most part, but it is quickly becoming less applicable. During the times when contract law was being formed in Europe, the identity of the parties was rarely in question. A buyer and a seller came together and agreed on the terms of a sale. The practice of reducing the agreement to writing became more common when the ancestors of today’s statutes of frauds were enacted. The point is that the parties knew and recognized each other.

It is true that the industrial revolution changed the way business was conducted. When most people lived in villages, it was easy to know everyone else. But when cities became the place where most of the population lived, it became nearly impossible to know everyone. However, it was still possible for a buyer to remember from whom he purchased the defective oven, for example. The initial interaction for the sale allowed a buyer to collect enough information about the seller to know whom to sue if a problem arose.

As more business is conducted online, knowing with whom you are dealing becomes more difficult. Strong encryption presents a new problem, which, in my opinion, serves to render the application of our current laws to online transactions impossible. Although strong encryption creates a system wide problem of enforcement, I narrow my discussion to contract law and the problems created there.

The Difficulty of Enforcing Contracts Online

As the use of public key encryption becomes more prevalent, traditional methods of enforcing contracts become obsolete. Anonymity undercuts many of the assumptions that the current system of law relies on, specifically in the areas of identifying the parties, determining their locations, and establishing the location of the transaction.

People are recognizable by their appearance. But online, people have no faces. With public key encryption, they do not even have names. Without knowing either the identity or the location of the parties, the current legal system cannot operate. No complaint can be filed, no process can be served, no jurisdiction can be established, no judgment can be rendered, and no judgment can be enforced. Even if a digital signature can be used to prove that the owner of a certain public key is really who he says he is, there is no way to link the real-space identity of this person to his cyber-space identity.

Allow me to illustrate the problem with an example. I am an editor. I advertise my services online. I am contacted via email by a certain entity that wishes to purchase my services. Both of us use public key encryption. I do not know who the other party is in real-space, and it does not know who I am in real-space. We come to an agreement as to the editing of a certain text, and we both digitally sign the agreement. Three months after the delivery of the finished product, I have yet to receive any payment from my employer. I have submitted numerous emails requesting status of payment, all of which have been ignored. What are my options? Traditionally, I would have brought suit against my employer for breach of contract. But since I do not know who my employer is, how do I sue it? I cannot go to its offices because I do not know where those offices are. Even if I file suit reserving the option to name a defendant in the future, I would not be able to enforce my remedy against it. Must I absorb the loss and give up doing business online to avoid future losses? The answer to this question is clearly no.

II. AN OLD SOLUTION TO A NEW PROBLEM

The problem of enforcing the obligations of parties is not a new one. In fact, before there were any formal laws governing contracts, non-legal means were used to enforce the obligations of parties. Historically, two broad forms of non-legal enforcement were used to solve a dispute: violence and reputational harm.

Violence, or the threat of violence, can be an effective means of controlling small groups of people. In feudal times, violence was not only an effective means of taking control, but also of keeping it. If you did not pay your rent, your landlord may send some of his thugs to remind you that payment is due. Governments also found violence to be effective in deterring crime. Public executions were not uncommon in medieval Europe. Going back even farther to the time of the Roman Empire, crucifixions were commonly employed. Violence is a common mob behavior used to steer societal norms. Take for example the infamous witch-hunts of early American history. The common belief was that if you are not part of the hunters, you are one of the hunted. People were eager to join the hunts so that they would not be branded witches and killed themselves. The reason violence was effective was because it was visible to the community, memorable due to the gruesomeness, and could be strongly associated with the badness of the person being punished.[5]

Reputational harm could also be employed, and in fact has become the preferred method of punishment. Gossip was especially effective in small communities where word could circulate quickly and the wrongdoer could be identified easily. The reason gossip was effective was that it communicated to other people the risk of doing any business with the wrongdoer. This was costly to the wrongdoer and gave him an incentive to correct his wrong and remain honest in the future. Gossip was a relatively mild form of enforcement.

If gossip failed to work, perhaps because the wrongdoer was an extreme deviant and cared little about the gossip, a harsher form of reputational injury could be applied. This harsher penalty is sometimes known as shaming.[6] Shaming is a much broader form of reputational harm as it exposes the wrongdoer to a much larger audience. Shaming was often seen in medieval Europe where a wrongdoer would be locked in a pillory. Anyone walking by could see the wrongdoer. Branding was also common. There is the well known example of the scarlet “A” worn by the adulteress in “The Scarlet Letter.” Anyone who would see the brand or the letter would know the offense of the person wearing it.

Modern Low-Tech Examples of Non-Legal Enforcement

Even today, there are instances where non-legal methods are employed to enforce obligations. Although today we have laws that govern most disputes that could arise between parties, some communities prefer to apply their own systems of law. Examples of this can be found in Shasta County and in the diamond and cotton industries. The conclusion to be drawn from these two examples is that non-legal enforcement mechanisms, especially reputational mechanisms, can be used effectively to enforce the obligations of parties without relying on a formal legal system.

Shasta County

In his book, “Order Without Law: How Neighbors Settle Disputes”, Robert C. Ellickson describes how the residents of Shasta County rely on social norms to work out disputes between themselves. The residents of Shasta County emphasize neighborliness through strong social norms. Residents of Shasta County fall into two general sub-groups: ranchers and ranchette owners.

Most of Shasta County’s ranchers raise cattle. Many of the disputes between residents arise directly from the problems associated with raising cattle. Ranchers of Shasta County belong to a trade organization called the Cattlemen’s Association. The Cattlemen’s Association is a nationwide industry group with its own set of rules. Membership in this group is important to the success of the rancher as many business relations form through the group. The group provides an exclusive market for information about market conditions, reputations of other members, and other social factors. Members of the Cattlemen’s Association share strong social norms that emphasize trustworthiness, honesty, reputational integrity, and social awareness.

Ranchette owners of Shasta County own small parcels of land. Unlike ranchers, ranchette owners may raise various animals as a hobby rather than to make a living. Ranchette owners do not belong to formal associations, but are instead bound by their common geography. There are cases documenting ranchette owners turning to the Cattlemen’s Association for assistance in solving a problem with a neighboring rancher. Ranchette owners, like most residents other residents, emphasize the importance of being a good neighbor and maintaining a good reputation in the community.

The focus of Ellickson’s book is the resolution of disputes that arise as a result of stray cattle including injury to property and trespass. Shasta County residents prefer the use of three non-legal methods before turning to the law for help.[7] First, residents apply self-help mechanisms such as gossip, violence or threat of violence. Next, they seek help from public officials such as the county supervisor or the animal control officer. Finally, they file informal claims for compensation without the help of an attorney. If all of the above fail, they file a formal lawsuit.

As Ellickson describes, “people tend to know one another, and they value their reputations in the community…Residents of the Northeastern Sector foothills seem quite conscious of the role of gossip in their system of social control.”[8] Reputation is highly valued in this community, and residents are strongly encouraged to employ reputation-based sanctions before resorting to legal sanctions.

In the majority of cases, this solution works. But in the event that it does not, violence or the threat of violence is used. The problem usually arises from one rancher’s cattle repeatedly trespassing or causing damage to another resident’s property. If the trespass is not corrected according to the norms of the community, residents have been known to make threats on the life of the trespassing cow. These threats are rarely carried out, as Ellickson notes, but there have been instances where a few cows have been shot.

If threat of violence proves ineffective, residents will often turn to private organizations for help: most often to the Cattlemen’s Association. The Cattlemen’s Association, which most if not all of the Shasta ranchers belong to, applies pressure on the wrongdoing member to correct his mistakes. This method is highly effective since it is more public and it comes from a group that is important to the rancher’s business. Bad reputation with the members of the Cattlemen’s Association is bad for business.

When all reputational methods fail, perhaps because the rancher cares little for his reputation, residents of Shasta County seek the help of public officials such as the animal control officer, brand inspector, or county supervisor. The animal control officer and brand inspector assist in the identification of the brand on the animal, and in the return of the animal to its owner. Repeat offenses sometimes lead residents to file complaints with the county supervisor in hopes of changing the ranging laws of their area. Such rezoning is seen as a drastic measure usually to be avoided. It is most often the ranchette owners who use this form of remedy, as they are less able to exert pressure through other means. [9]

On rare occasions, residents will submit formal claims for compensation to the wrongdoing party. The more accepted form of accounting is what Ellickson terms “informal accounting.” With informal accounting, a resident is allowed to keep track of the costs he incurs as a result of his neighbor’s wrongdoing, and offset these costs with any damage that his wrongdoing may have caused the other. Balances are also offset by other events such as the neighbor fixing a shared fence, or caring for a stray cow until it is recovered. If the accounts balance out, then there is no problem. As one of the cattlemen interviewed by Ellickson said, “I figure it will balance out in the long run.”[10]

There is a clear norm against seeking monetary damages. But if the informal accounts do not balance, a rancher may file a formal claim. Lawyers are never consulted for this purpose because the residents believe that when a lawyer gets involved, the only person who makes money is the lawyer - another indication of the norm against employing a formal legal system. The residents strongly believe that being a good neighbor means not suing other neighbors. However, if all of the above means fail, the residents file a formal legal complaint.

The Diamond Industry

In “Opting Out of the Legal System: Extralegal Contractual Relations in the Diamond Industry”, Lisa Bernstein describes how members of the diamond industry resort to their own system of rules to handle disputes among themselves. Bernstein describes the operations of the New York Diamond Dealers Club, more commonly known as the bourse. The bourse is composed of sight holders, manufacturers, wholesalers, and brokers. Overall, the bourse has over 2,000 members with many more on a waiting list.[11] The New York Bourse is itself a member of an international group of bourses known as the World Federation of Diamond Bourses (WFDB). Like all the other bourses, the New York bourse is an informal market where members come together to do business. The bourse applies its own set of rules that control virtually every aspect of doing business in the diamond industry. These rules reinforce the social norms adopted by the members of the bourse that encourage fair and honest dealings. The keys to the success of the bourse are the use of arbitration along with the consequences accompanying failure to comply with arbitration, and the value placed on reputation by members of the group.

In the diamond industry, arbitration is the mandatory method of settling disputes. In fact, agreement to arbitrate is a prerequisite for membership in the bourse.[12] Since much of the information that is discussed during arbitration is related to the business of the parties, all of the records of the arbitration remain secret.[13] Any member who files a formal legal complaint against another member is subject to suspension or expulsion. Either one of these penalties is fatal to a diamond dealer as very little business is conducted with non-members. If a member fails to comply with the terms of the arbitration, his picture and the decision of the arbitrator are posted on the bulletin boards of all the Bourses in the world.[14] This indicates to all the other members that this person is not to be trusted or dealt with. In the diamond industry, this type of reputational injury is fatal.

Diamond industry insiders rely heavily on the reputation of other members. Reputation conveys important information about the likelihood of a sound business transaction with any given person. Members value their reputations because of the inherent economic value of reputation. Good reputation translates into good business. Bad reputation means no business. Since most members of the diamond industry rely on social norms when doing business, reputation offers a quick indicator of the success of the transaction and of the trustworthiness of the other party.

The Cotton Industry

In “Private Commercial Law in the Cotton Industry: Creating Cooperation Through Rules, Norms, and Institutions,” Lisa Bernstein examines the way in which the cotton industry in the United States uses a private system of norms to control dealings in the industry. The cotton industry is divided into two groups: merchants and mills. Each group may be composed of several trade associations. Most merchant-to-mill transactions are controlled by the Southern Mill Rules, which is a standard set of rules adopted by both the miller and merchant associations.[15] These rules, which are similar to other such rules in the industry, require members to settle disputes using the arbitration services provided by the industry groups.[16] Failure to comply with the findings of the arbitrator is grounds for expulsion from the group.[17] Expulsions are widely publicized and tend to have a detrimental affect on the business of the expelled party.

Members of the cotton industry rely heavily on the reputation of other members when conducting their business. Much of the business in the cotton industry is conducted informally with a telephone call rather than with a contract. [18] This type of dealing requires that the parties trust each other and have confidence in the good faith of the other. As Bernstein writes, “individual transactors attach great weight to the reputation of their transacting partners, and the institutions that create and administer the [private legal system] also engage in many activities that are designed to strengthen the social basis of trade and to promote the flow of reputation-related information though the industry."[19] Because of the form that business transactions take in the cotton industry, members often prefer to get to know each other socially before they conduct business. For this reason, the cotton industry groups often host social events for their members. It is not uncommon for potential business partners to go fishing, hunting, or have dinner at each other’s homes before they conduct any business.

Why Non-Legal Enforcement Works

The above examples demonstrate how complex real-space systems can operate successfully without relying on formal legal rules. Even though it seems that Shasta County cattlemen have little in common with diamond dealers, or that cotton millers have little in common with diamond wholesalers, each of these groups has been able to sustain its respective non-legal system of rules for many years. In fact, the cotton industry has been able to maintain its informal system of laws since the mid-1800’s, surviving substantial changes in the industry during that time.[20]

Despite the marked differences between each of the respective groups, they all possess three qualities that in my opinion are directly responsible for the respective success of each group. First, each system relies on membership in common group. Second, each member of the groups anticipates continued and repeating relationships with other members of the group. Finally, each group places high value on the reputation of other members. I will refer to these elements as membership, relationship, and reputation, respectively.

Membership

The first factor that all of the systems have in common is membership in a group. In my opinion, group affiliation is the foundation for the success of a non-legal system of enforcement. The structure of the groups offers an exclusive, closely regulated, safe environment to conduct business. It is a place where members can come together to do business knowing that the transactions in which they engage are governed by a system of norms that they have selected. Membership in the group is a sign of trustworthiness. There are four reasons why membership is important to the success of a non-legal system. First, membership creates a place for the exchange of information about other members. Second, membership provides a place to settle disputes without relying on a formal legal system. Third, membership offers financial benefits to members. Finally, membership serves an important social function.

Information Exchange

The first important function of group membership is that of an information exchange. Information flow is vital for the success of the group because it minimizes the risks of doing business and it encourages members to remain honest. The more reliable the information is, the less likely it is that disputes will arise. Information becomes more reliable when other members can verify it. Such verification can occur when dealings are continued and repeated over a long period of time. The more the members deal with each other, the more likely that the true business nature of the members will be revealed. Fair dealings produce good information and increase business. Similarly, unfair dealings produce bad information and a decrease in business.

The subject matter of the circulated information generally relates to the reputation of other members. Membership in and of itself signals a member’s willingness to transact honestly and fairly.[21] This is especially important in the diamond industry for small or new dealers who have not had an opportunity to develop their reputations.[22] Members share information about other members regarding success or failures in transactions. Also, members are able to post claims about gossip about them that is false or baseless.[23] This is truer in the diamond and cotton industries where the groups are better defined than in Shasta County. However, even in Shasta County, complex systems of gossip spread information about the behavior of one resident.

According to Bernstein, good reputation in the cotton industry develops from a member’s record for paying his debts on time, meeting contractual obligations, willingness to renegotiate commitments when conditions have changed, record for one-time delivery, and delivery of the quality of goods specified.[24] Good reputation is developed in the diamond industry in much the same way as it does in the cotton industry. Shasta County residents rely less on business transaction information and more on the neighborly behavior of members. In all of these cases, this information helps establish the reputation of members, and other members rely on this information.

The information that is communicated in these groups is not limited to that of reputation. Often, information about market conditions and other industry related topics is shared. Bernstein noted that one retired member of the diamond industry, who was granted honorary status, continued to cooperate in the bourse to keep up with market conditions. Apart from the business reputation of members, social reputations seem to be of value to the cotton industry and the residents of Shasta County. As Bernstein notes, a member’s social reputation is intertwined with his business reputation.[25] Since social reputation is important to a healthy business, this type of information is also communicated. As one cotton industry participant said, “[s]o you tell me, when you want to do business, who will you call, the guy you like or the guy you don’t like?”[26] Shasta County residents rely on social reputation to achieve certain objectives such as repairing common fences or borrowing the help of neighbors. The Shasta group strongly encourages neighborliness. When one neighbor fails to meet the standards expected of him, negative gossip will begin to circulate until the desired result is achieved.[27] It is clear that in many cases social reputation is as important as business reputation.

The content of the information is transmitted in a variety of ways, usually depending on the nature of the information. Gossip seems to be the prevalent means of transferring information. The group environment is highly conducive to gossip because the coming-together of the members in one place allows gossip to spread quickly and efficiently. It is much easier to aggregate information about players in the industry if they all belong to the same group. Gossip is spread in social gatherings or between private circles within each of the groups. The subject matter of the gossip is obviously not intended to remain secret. If it was, the desired effect of resolving the problem could not be achieved. The cotton industry goes as far as recording such information and making it available to other members through a request placed with the Secretary of the group.[28]

Gossip works well for communicating information about a member’s reputation. There are other means used by the groups to transmit more formal information about things such as changes in the rules regulating certain transactions or failure of a member to comply with an arbitration order. For example, the cotton industry regularly publishes circulars describing changes in the group’s rules as well as clarifications in the trade rules to avoid confusion.[29] The diamond industry uses a different method for publicizing a member’s failure to comply with an arbitration ruling. In such a case, a picture of the member, along with the ruling from the arbitration proceeding are posted on the bulletin boards of every bourse in the world.[30] This is an effective way of communicating to all other members that this particular member is not to be trusted. Most often, however, gossip spreads this information before the news is even posted.

Dispute Settlement

A second factor of membership that is important to the success of a non-legal system is that membership encourages rapid dispute resolution. Membership in a group makes arbitration available to the disputants. Arbitration is mandatory in the cotton and diamond industries for good reasons. Arbitration is cheaper and more efficient than litigation. Cotton groups require a payment of $400 from each party, which is refunded to the prevailing party.[31] This is also true in the diamond industry where disputing parties are required to pay an arbitration fee that is refunded to the prevailing party.[32] As Bernstein notes, “[arbitration] encourages settlement by making arbitral outcomes easier to predicts and promotes flexibility in the course of contracting relationships by refusing to transform extralegal understandings or patterns of adjustments into legally enforceable obligations.”[33] The social norms employed by members of the groups provide a basis for contract obligations that is clearly understood and agreed to by both parties, but that is not necessarily captured in the terms of the contract. The risk that such understandings will not be taken into account by a court encourages members to settle their disputes within the governing rules of their industry. Arbitration is not required in Shasta County, but it is well understood that good neighbors do not sue each other. Disputes are settled informally. Residents usually resort to gossip or self-help methods such as negative gossip, or in the somewhat more extreme cases, violence against cattle, to resolve their disputes.[34]

Apart from the procedures that encourage dispute settlement, reputation-based incentives also encourage quick dispute resolution. It is in the best interest of both parties to resolve disputes quickly and quietly so that neither party’s reputation is injured. Since information travels quickly in these systems, both parties have an incentive to keep the dispute as confidential and cooperative as possible. As discussed above, the reputation of a participant in the cotton industry is partly determined by his willingness to renegotiate terms when circumstances have changed. Renegotiations of such terms helps avoid conflict up-front while at the same time signaling a member’s willingness to avoid such conflict. These signals increase the value of a member's reputation. Good reputation is good for business, no matter which system is involved. As Bernstein notes "many transactions [in the diamond industry] are still consummated on the basis of trust and truthfulness, this is done because these qualities are viewed as good for business, a way to make a profit."[35] Reputation is directly linked to the financial benefits of membership. Although not obvious, the residents of Shasta County also benefit financially from a good reputation. Good neighbors take care of their common fence, splitting costs or bearing all the cost if necessary [36]; good neighbors help you clear your field; good neighbors take care of your stray cattle when it wanders onto their land.[37]

Financial Benefits

Third, membership is important to the financial success of a member. These types of financial benefits are important to the overall success of the system because they encourage members to remain in the group and to follow the rules of the group. The bourse gives its members a strong competitive advantage over non-members in the diamond industry. Since 80% of the world's diamonds flow through the hands of bourse members, it is vital that one have access to this market. Membership in Shasta’s geographic group encourages neighborly behavior and the sharing of common costs. There is a universal understanding among the members of the groups that the group benefits everyone. As one cotton industry insider noted about why people gossip about wrongdoers, “no on wants to see the industry hurt by an unscrupulous person.”[38] The norms established by the group reward honesty because it benefits everyone. Conversely, dishonesty is punished because it hurts the group.

Most business transactions in these groups are conducted between members. The bourse most clearly illustrates this point. The bourse is a market for diamonds apart from simply being an industry organization. Diamond industry participants come to the Bourse to buy and sell diamonds. Access to a steady supply of diamonds is vital to the profitable operation in this industry.[39] As mentioned above, getting access to diamonds requires membership in the bourse. Without such membership, little business can be done. The economic benefits of membership in the bourse are also evidenced by the fact that membership is highly sought after by most market participants.[40] The cotton industry also nicely illustrates this point. As Bernstein writes, “membership in a shippers association strongly affects the profitability of a merchant’s domestic business.” [41]

Social Function

The final factor of group membership that contributes to the overall success of non-legal systems is the social function that groups serve. Most of the groups host social events such as dances, golf competitions, and dinners. One group in the cotton industry, for example, sponsors the local debutante ball and a well-known domino tournament.[42] A domino tournament may seem like an odd event to sponsor, but considering the fact that much of the business in the cotton industry is based on social relationships, and that social relationships develop at such an event, it makes perfect sense that such relationships are fostered and encouraged.

Residents of Shasta County hold similar social events including large parties and dinners. Besides being a perfect forum for gossip, these events help build a sense of community and cooperation. A sense of belonging, apart from a business aspect, makes group membership more enjoyable. The more enjoyable the group is, the more likely members will remain in it. As social relationships grow stronger between members in the group, business transactions become more secure, and informal dispute resolution becomes favored over legal action. This is true for all the groups. As one member of the cotton industry commented, “[m]erchants take mill buyers on hunting trips just like in any other business…In the process, relationships develop. Over time, a buyer gets the idea that he wants to deal with me not just because of our business relationship, but also because of our personal relationship.”[43] The bonds that develop as a result of group membership clearly add to the success of the non-legal system.

Relationship

The next element that is critical to the success of the non-legal system is the expectation of all of the members that contact with other members will be continuing and repeating. In other words, each member expects that he will deal with another member many times during his membership. The expectation that a relationship will be continuing is important to the success of the non-legal system in that it discourages disputes and that it provides economic benefits to the members, thereby giving an incentive to keep relations on good terms.

Discourage Disputes

The expectation that relationships will continue over a long period of time discourages disputes. If parties anticipated repeated contact, they want to at least ensure that that contact is pleasant. Besides the economic advantage of keeping relationships pleasant, which I discuss below, there is a fundamental social advantage to doing so. Shasta County provides the best example of this. It is to the advantage of all neighbors to stay on good terms with the other neighbors so that living near someone does not become uncomfortable. Apart from quality of life considerations, there are other non-economic advantages to being on good terms with your neighbor. For example, your neighbor can watch your house for you while you are away. Shasta’s social norms in particular are well adapted to minimizing disputes between members. According to Ellickson, residents employ a type of informal accounting among themselves. When one member does a favor for the other, the first expects an offsetting favor in the future – to balance the accounts so-to-speak.[44] Disputes do not arise immediately if the accounts do not balance. As one resident said, “I figure it will balance out in the long run…I hope they’ll do the same for me.”[45] This method of informal accounting grants some leniency for the wrongs of a member. The norms inherently build a delay mechanism into the system that allows for problems to work themselves out, thereby reducing disputes.

But this type of benefit is not limited to Shasta County. Leniency is important in the cotton industry as well. Members are encouraged to renegotiate terms of a contract that may have changed as a result of unforeseen circumstances. This process reduces the number of disputes in the industry. In the long run, it is beneficial for one party to renegotiate because the next time the parties contract, the tables may be turned. The social bonds between members minimize misunderstandings, which reduces the number of disputes, because the parties know each other on a more personal level and understand the way the other thinks. Also, since their social reputations are tied with their business reputations, it is to their advantage to be lenient so that both reputations are protected.

Economic Benefits

The more obvious incentive for members to maintain their relationships with other members is that good relationships mean good business. There is a direct correlation between the reputation of a member and the success of his business. It is universally true among the groups examined that members prefer to do business with other members with whom they have good relations.

The reliance on good relationships is prevalent in the cotton industry. Members of this group build social bonds before they begin doing business. Business is conducted on a more personal level, which further encourages honesty and fair dealings. As was mentioned above, one member of the industry said that he would rather do business with someone he knows personally than with someone he does not.[46] Once again, membership is vital to the realization of this expectation. Without the group, contact with other members of the industry would be more difficult, and thus business would be inherently more risky. Without the social relationships, there is no alternative non-legal means to enforce a contract or to avoid dispute. The expectation that business with members will be continued and repeating has been proven reasonable again and again over the long span of the industry’s history.

The residents of Shasta County obtain a different kind of financial incentive from maintaining good relationships with each other. Besides the fact that neighbors live near each other and have to see each other frequently, it is mutually beneficial for the neighbors to act this way: costs of building or repairing fences can be shared; equipment can be borrowed to help clear your field or dig a trench; your neighbor can watch your property while you are away. These are all ways in which neighbors reduce their costs. The savings encourage good relationships.

For those residents who are members of the Cattlemen’s Association, good relationships are beneficial in many of the same ways as they are in the cotton industry. Business is conducted on a personal level after the members have had a chance to get to know each other personally. The Cattlemen’s Association sponsors social event for such reasons. Members of the Cattlemen’s Association frequently do business with each other and they expect to do so in the future. This expectation has in fact been realized for the many years that cattlemen have owned land in Shasta County.

Diamond industry participants have much to gain from strong relationships. Relationships are the main factor determining who gets to purchase a certain box of rough diamonds and for how much. Good relationships are rewarded with a better selection of diamonds and more favorable prices. And as is true in all of the systems, good relationships produce good reputations. Good reputations, in turn, produce better business opportunities in the future. It is understood that personal relationships are important to the success of the respective group. Such relationships necessarily give rise to reputation-based information, which is heavily relied upon by other members.

Reputation

The final factor contributing to the success of the non-legal systems is the high value that members place on reputation. Reputation serves as a signal to other members that one deals fairly and honestly. Members rely on reputation as a basis for many of the business decisions they make. Maintaining an emphasis on the importance of reputation is crucial to the success of the non-legal system in three ways. First, it provides economic benefits to its members. Second, it encourages fair dealings between members. Finally, and as a result of the two prior benefits, putting value on reputation helps the groups avoid conflict.

Economic Benefits

Good reputation is good for business no matter which way you look at it. Reputation is the foundation on which most of the transactions in the group are based on. Members rely on the reputation of other members as a signal for the likelihood of success of any given transaction. It is for this key reason that much of the information communicated in the group revolves around reputation.

It is clearly understood that good reputation is good for business. As one cotton insider said, “[r]eputation is essential in this business. Millions of dollars of business will be done on the basis of a thirty-second phone call.”[47] And another stated, “Reputation matters because you conclude sales one day, and the confirmations do not follow for as much as a week or ten days, so during that time period, the integrity of a person’s word is the only thing that bonds the contract.”[48] Good reputation increases business because other members can trust that deliveries will be made on time, that the quality will be as required, or that payment will be made on time. The better a member’s reputation, the more likely other members will be willing to deal with him.

Diamond traders conduct their business as informally as cotton dealers, often entering contracts with a handshake and an utterance of “Mazel and Broche.”[49] Good reputation is vital to the success of any party in the diamond industry for most, if not all, of the same reasons that apply in the cotton industry.

But reputation does not only have an effect inside the particular groups. This is particularly true in the cotton industry where banks rely heavily on the reputation of their potential lenders when making loan decisions. As Bernstein noted, “Banks inquire directly into the commercial reputation of their borrowers, and a merchant’s reputation, along with the reputation of the mills he sells to, strongly affects the interest rate he will have to pay.”[50] Clearly, reputation has much to do with the success of any group member, even outside of the group. It is in the best interest of all the parties to keep as good a reputation as possible so that financing will be available when necessary.

But the financial benefits of keeping a good reputation are not limited to the business context. In Shasta County, a good reputation among one’s neighbors leads to the sharing of costs relating to fencing, or the care of stray cattle. Neighbors with good reputations are more liked in the community and are treated with more respect. As Ellickson noted, some families in the Shasta area have lived there for many years, and are intent on staying there for many more. These families in particular, seem intent to maintain their good reputation.[51]

Good reputation has benefits outside of the business circles. The interweaving of social and business reputation aids the groups to maintain order and to encourage fair dealings. If a member’s overall economic success is tied to the way he acts socially as well as in the business context, he will be encouraged to maintain good relations with other members, thereby maintaining a good reputation.

Encourage Fair Dealing

The emphasis that is placed on good reputation, and the economic benefits that result from maintaining a good reputation encourage members to deal fairly. Since financial success is tied directly to a member’s reputation, actions that increase the value of that member’s reputation will be preferred. Actions that increase a member’s reputation include dealing fairly. The group as a whole benefits from fair dealings. Through the mechanism of reputation, cooperation is encouraged and opportunism is discouraged. It is understood that good reputations develop from helping other members because in the future, the tables may turn and the reciprocal situation may arise. As Bernstein noted, “Because…non-cooperative responses tend to reduce his future trading opportunities, the long-run costs of defection will often be greater than the short-term gain from defection, so a transactor who is not in financial distress is more likely to cooperate than defect.”[52] In the long-term, the reputational penalty from failing to deal fairly is so costly to the member’s business that obligations are transformed into a self-controlling way of encouraging cooperation. Members benefit more if they cooperate.

This is also true outside of the business context. Shasta County residents derive a different type of benefit from maintaining good reputations. For the same reasons that it would be beneficial for a cotton dealer to maintain a good reputation, Shasta County residents also have incentive to maintain the good reputations. As was mentioned above, some families have lived in Shasta County for many years. They have an expectation that they will remain there for many more. This expectation will naturally follow by the expectation that contact with members in the community will be repeating and continuing. In order to maintain their good reputation with the community, these families have an incentive to be neighborly, i.e. deal fairly with other residents.

Avoid Conflict

Since reputation is highly valued, and fair dealings increase the value of one’s reputation, members have an incentive to minimize conflicts so that their respective reputations will be spared. Some dispute seems inevitable, but there is a fine line between some disputes and disputes that seem to recur frequently. The reputation of a member who seems to be in dispute frequently will be damaged. Other members will be hesitant to initiate business with that member for the fear that they too may be drawn into a dispute. It seems inherent that the structure of the system discourages conflicts. For example, when the circumstances of a cotton deal change, and one party requires renegotiation, the other party has an incentive to renegotiate because his reputation increases in value. The act of helping the other party is seen as fair dealing and is rewarded in terms of reputation-based incentives.

In Shasta County, the reputation of a resident is measured by his neighborliness. The informal system of accounting that is used by members inherently encourages cooperation rather than conflict. When one resident allows his neighbor to borrow his tractor for example, this is seen as a neighborly gesture, which is rewarded with an increase in the value of that resident’s reputation. Although the informal accounting records are not balanced, it is understood that the willingness of neighbors to allow such imbalances is the proper way to act in the community. It is also understood that these imbalances will be corrected in the future and that a member who does not correct such imbalances may incur injury to his reputation in the form of negative gossip. The elements of membership, the circulation of information, and the emphasis on reputation allow non-legal systems to minimize disputes and to function efficiently.

III. REPUTATION AND STRONG PRIVACY

The Framework

The above discussion demonstrates how the diamond industry, the cotton industry and the residents of Shasta County are able to enforce the obligations of their members without relying on formal legal rules. The success these groups have achieved in implementing their non-legal systems of enforcing obligations is due in large part to three factors. The first of these factors is membership in a common group. Membership allows for communication about transactions, provides a mechanism for resolving disputes, provides financial benefits, and serves a social function. The second factor is the anticipation of all the members that interactions with other members is likely to be repeating and continuing, which results in the incentive to build and maintain good relations with one another. The final factor is the overall importance that reputation plays in the decision processes of the members.

These three factors taken together present a framework for building a non-legal system to enforce obligations of parties in a world of strong privacy. To briefly restate the problem, how can parties enforce each other’s contractual obligations when neither knows whom the other is, thereby rendering traditional legal remedies, such as a lawsuit, useless? The framework succeeds in solving this problem by eliminating the reliance on knowing the real-space identity of the parties, and by creating an alternate, non-legal system for dealing with disputes that is based on reputation.

Eliminating Reliance on Real-Space Identities

As discussed above, the current legal system requires that the identities of the parties be known for jurisdictional and enforcement purposes. This reliance eliminates the possibility of applying the system of law to a situation where parties regularly use public key encryption. The framework does away with this requirement by on the one hand, relying on the way in which public key encryption works, and on the other, relying on reputation-based mechanisms.

Public key encryption allows a party to remain anonymous and at the same time prove its identity. This means that although parties cannot know each other’s real-space identity, either can prove his cyber-space identity by decrypting a message encrypted in his public key. The ability to tie a certain public key to a certain online identity provides the foundation for the solution. Online entities are able to maintain their unique identity, which means that reputation can be attributed to these identities. Without this feature there would be no way to enforce reputation-based measures. Being able to point out a certain online identity allows for reputation-based mechanisms to have a meaningful effect on the way parties transact online.

The fact that public key encryption allows entities to "identify" themselves dispels the argument that a party who wishes to avoid an obligation could simply create another cyber-space identity by rolling out a new public-private key pair. This may be true in a few circumstances where a party does not care to establish a good reputation. However, such behavior will likely damage that party’s business in the long run. Since reputation is tied to a certain public key, not having a reputation attached to that key makes that party suspect. Is this a new business or is it seeking to avoid something? Not wanting to establish a good reputation, or having a bad reputation, signals to other members that this party is not to be trusted. Also, abandoning a certain public key is sends a negative signal to other parties. Since there is a cost associated with building reputation, it is less likely that entities will be quick to abandon one reputation and build another.

The framework establishes an economic incentive for members to maintain good reputations. Since reputation is the only way parties can predict the success of a given transaction, it is costly for parties to create new identities each time they transact. New identities mean new reputations, and reputation has to be developed. Developing a reputation takes time and money. As long as it remains costly to build up one’s reputation, parties will not be so quick to change identities.

Reliance on Reputation

Strong privacy eliminates the possibility of determining the real-space identity of another party in order to use conventional legal solutions, such as suing, to resolve disputes with that party. If such legal solutions are no longer available, how can parties enforce the contractual obligations of others? The solution is that reputation can be used to signal the likelihood of the success of a transaction before a transaction is entered into.

In the cotton and diamond industries, as well as in Shasta County, reputation plays a vital role in inter-member dealings. Reputation can be used online to minimize the likelihood of disputes in the same manner that reputation serves as a signal to members in the diamond or cotton industries. The membership and relationship prongs of the framework are required elements if reputation is to achieve the maximum effect of deterring dishonest business practices, which thereby minimizes disputes.

Membership in a common group serves the crucial role of allowing for the aggregation and communication of reputation-based information. Without a group, it would be difficult for parties to obtain information about any other party. In the same manner that groups in the cotton industry aggregate data on the reputation of their members, an online group of buyers could aid in collecting information about sellers, for example. Membership encourages communication because the group brings together individuals with common interests. Members gain a bargaining power in the same way that a union gives laborers bargaining power.

It is likely that a certain party will be a member of different groups, and the critic may argue that membership in one group may not allow members of another group to access information about a party that is a member of both. In other words, a party may have good reputation in one group, but a bad reputation in another and that information may not be available to members of the other group. However, the framework is designed in such a way as to give a party the incentive to maximize his good reputation no matter what group he is in. As was discussed above, there are many financial incentives to maintaining a good reputation.

Also, the reputation of one party in one group is wholly irrelevant to members of another group. What matters to members within a group is the reputation of another member in that particular group. For example, assume a party is a member of both the cotton and diamond industries. The reputation of that party in the cotton industry is mutually exclusive of his reputation in the diamond industry when he is dealing with members of the cotton industry. All that the members of the cotton industry care about is whether that party has a good reputation in the cotton industry. Therefore, the flow of information between different groups is irrelevant in this system because the reputation of members is only important in the context of any given group.

Another important aspect of the reputation-based system of enforcement is that within each group the likelihood of a continued relationship is higher than it is in a system without distinct groups. Groups aggregate members as well as information. Financially speaking, sellers of DVD players would much rather have groups of DVD buyers to target. With the groups, such a targeted market exists. Therefore, both buyers and sellers can anticipate that as long as they are members of a particular group, business will be drawn to that group and the chance that the same parties will interact again in the future increases. It can also be assumed that if such repeating business is anticipated, then both parties will want to maintain good relationships with one another. Sellers have an incentive to treat their customers well so that they can make repeat purchases or perhaps recommend the seller to other buyers. Buyers have an incentive to deal fairly with sellers so that they have the opportunity to purchase in the future on perhaps better terms than someone who has never before dealt with that seller does.

The reputation-based framework is therefore an elegant way of dealing with the problem of enforcing obligations when strong privacy makes current legal systems useless. By emphasizing the benefits of groups and the benefits derived from maintaining a good reputation, the framework substitutes formal legal rules with non-legal solutions for avoiding disputes rather than attempting to solve them after the fact.

Modern Examples

No single organization has implemented all three elements of this framework as of yet. Perhaps the reason for incomplete implementation is the fact that strong privacy not a widely used method of doing business. However, of those organizations that have implemented some of the elements, three in particular deserve mentioning. Those three organizations are eBay, RepCheck, and BizRate.

These organizations have been testing portions of the framework in the real world, or in cyber-space, as may be appropriate. Therefore, they provide for an interesting foreshadowing of the overall success of the framework when strong privacy becomes widely used, not only by businesses but also by the average person.

eBay

eBay is an online auction company. The core of eBay's business is that it provides buyers and sellers of goods a market to interact. eBay is the online mix of a public auction and a flea market. By far, eBay is the most successful online auction site with millions of users and millions of items for sale.[53] eBay is also one of the few Internet companies to reach profitability[54], and maintain a strong market valuation.[55]

eBay is the best example of a current Internet company that takes advantage of the framework.

However, eBay's use of the framework is not complete. For example, eBay is not able to achieve the expectation that contact between members will be continuing and repeating. Also, the use of screen-names does not achieve the same level of privacy that public key encryption does. Nevertheless, eBay's model is still the best example of an organization that has successfully implemented portions of the framework. Its model has more positives than negatives, which is more than could be said about other such organizations.

The first of these positives is that eBay has successfully implemented most of the membership element. eBay users are known for their sense of community. eBay emphasizes community values such as "We believe everyone has something to contribute," and "We believe that an honest, open environment can bring out the best in people."[56] A key to the success of eBay is the fact that its members harbor a sense of unity and affiliation. In many of the same ways that cotton dealers, diamond traders, and Shasta County residents feel a sense of belonging, eBay members belong to a unique group – granted, the eBay group is different in that it is completely online.

The fact that the group is online means communication between members is even easier than it is for real-world groups. Information is exchanged via email or as comments posted directly on the site. The ability of members to comment on their transactions with other members increases the value of being a member. After winning a bid for an item, buyers are given the opportunity to comment on the transaction as a whole. Buyers are asked to comment directly on their dealings with the seller, the condition of the item as compared to the way it was represented by the seller, how quickly the item was delivered, and many other aspects of the business transaction.[57] All of these elements, along with an overall rating of the transaction, are recorded and displayed with the seller's profile. This information is available to anyone who wants it regardless of whether he is a buyer or simply browsing.

Also, membership in the eBay group creates economic benefits. This is clear from the recent news of large companies such as IBM selling their goods on eBay.[58] The sheer volume of users accessing the wide variety of products gives sellers the opportunity to reach a market larger than they could ever reach without such a service. The cost of reaching such a large market is also dramatically low compared to the real-world costs of reaching the same number of people – eBay only charges commissions on sales.

The second positive feature is that eBay has been successful in developing a method for reinforcing and contributing its members’ heavy reliance on reputation when making purchasing decisions. The reputation of sellers is very important to their financial success. The inherent risk of doing business online with someone whose identity is kept secret requires members to seek out other methods of ensuring the success of transaction. Buyers will therefore seek sellers with good reputations. Sellers have an incentive to maximize their reputation so that they can attract more business. Sellers increase the value of their reputations by being honest. The system encourages fair dealing in much the same way that the cotton and diamond industries encourage fair dealing. Financial incentives ensure that business transaction remain honest.

The final positive of eBay’s model is that it has implemented a method of resolving disputes between members. It is logical to assume that as the number of members increases, the number of disputes will also increase. Therefore, along with a fraud protection and prevention program called SafeHarbor[59], eBay provides for optional arbitration between disputing members through an independent organization named SquareTrade.[60] The SafeHarbor program requires that buyers take some preliminary steps such as attempting to communicate with the seller, or try to settle the dispute through SquareTrade, before the program applies. But if a member qualifies, eBay will reimburse that member for up to $175 of the amount that he lost.[61]

eBay's implementation of the three pronged framework is incomplete in that it does not foster the relationship prong. It is difficult, in a group of this size, to establish the expectation that interactions between the same buyer and seller will be continuing and repeating. A simple search for virtually any item, except the bigger ticket items, will reveal the multitudes of sellers from which the buyer may choose. A buyer can discriminate between sellers on a number of levels including reputation and price. This means that buyers are not tied in any way to one particular seller. Unlike the cotton industry, where business is done on a more personal level, there is no reason for a buyer to keep buying from the same seller if the buyer can get what he wants for cheaper somewhere else. The number of sellers and buyer on eBay influence the relationships that develop between members. This is perhaps the reason eBay emphasizes reliance on reputation rather than on relationship.

eBay's implementation of the framework is not complete in that it does not harbor the development of long-term relationships between its members. However, eBay is by far the best example of the framework in action today. Another example, although a less complete one, is Bizrate.

Bizrate

Bizrate is an online consumer rating company. Bizrate collects information directly from customers of online companies. Bizrate knows which customers bought from which business because the businesses make this information available to it. As a member of Bizrate, I receive email questionnaires about the quality of my experience with a particular business after I have placed an order with that business. This information is then aggregated, summarized, and posted on Bizrate’s web site for everyone to read. The reputation of online businesses is therefore calculated from the opinions of its customers.

Bizrate is a somewhat less appropriate example than eBay, but it is a good example nevertheless because it demonstrates that sellers value their reputation and that Bizrate encourages sellers to maintain good reputations. Bizrate forms a market for buyers and sellers, although the option to buy goods through Bizrate is a new feature. The market for sellers is somewhat more restricted as sellers must pay Bizrate for its services. Also, buyers can purchase through any listed seller, but do not have to be members of Bizrate. Only members of Bizrate can communicate their opinions, and then only if they choose to. The communication function of membership is fulfilled in the direction of buyers rating sellers, but not vice versa.

Members of Bizrate who are sellers value their reputations. Bizrate provides a good mechanism for capturing and communicating that reputation to other potential buyers. The proof that members value their reputations lies in the fact that they are willing to pay for Bizrate to perform such services for them. Bizrate provides online businesses, which allow Bizrate to continually survey all of the business’s customers, with a gold star logo.[62] Having such a logo displayed on web site signals to potential customers that the business is willing to listen to its customers’ opinions and seeks to improve its service. Bizrate accumulates customer responses to its surveys and generates an overall rating of a particular business based on those responses. The rating of the business is posted on Bizrate’s site. According to Bizrate, the rating is based on ten factors including customer support, ease of ordering, and price.[63] The rating is a good indicator of a business’s reputation – the higher the rating, the more likely the transaction will be successful. As the Internet community accepts Bizrate’s methods as the industry standard for measuring reputations online, sellers who can display this logo and are able to maintain a high rating will be more likely to draw business. Financial incentives to maintain good reputations, for sellers at least, is therefore met.

However, Bizrate falls short on a few aspects of the framework including some elements of membership and relationship.

The membership factor of the framework is not completely met in that information is not freely exchanged. In order to achieve a numerical rating, buyers are required to answer surveys based on scales. There is no room for writing out opinions because these cannot be easily translated into a comparable indicator. Even if there is room for such comments, the surveys themselves are aggregated. It is not possible to see the survey results by individual, which makes it more difficult for other members to respond to them. Free information flow requires that members be able to comment on each other’s opinions. Through such a process, it is possible to ascertain the true reputation of a member. Numerical representations are a useful first impression, but it is impossible to determine why a particular member responded in the way he did. Perhaps one customer is unhappy with the product selection and therefore rated the business poorly. Another customer, however, may not care about the product selection but care about prices. Without being able to see why a particular member responded the way he did, it is impossible to judge whether that response is based on grounds important to other members.

The relationship factor is also not met for the same reason it is not met by eBay – competition between members. Since sellers aggregate around the buyers, it is possible for the buyers to be selective. What more, since Bizrate only lists online businesses, geographical competitive advantages are eliminated. The ability to purchase goods online eliminates reliance on businesses that are closely located to the buyer, thereby increasing the selectivity of those buyers. Also, as mentioned before, knowing only the numerical rating of a business makes it difficult to predict the likelihood of success of the transaction because the particular factor that made one customer rate the business poorly may not have any relevance for this buyer. Therefore, it is difficult to encourage long-term relationships between members especially given the fact that Bizrate only rates online businesses, which have been less stable than brick and mortar businesses. Turnover in the Internet industry is high, therefore the difficulty of establishing long-term relationships is also high.

RepCheck

The final example of a current business using part of the framework is RepCheck. This company is the real-space version of Bizrate in that it accumulates and sells information about the reputations of businesses, but it differs from eBay and Bizrate in that it sells information about individuals as well. RepCheck accumulates information about people’s reputation from member feedback.[64] It is not clear from the description on the site what kind of information is collected and how. One thing is clear, however – we need to watch our backs. As part of the benefits of its service, RepCheck suggests that employers may want to learn about social reputation of their new employee or learn about the person that you are about to go on a date with. [65] The concept is interesting in that it expands outside of a strict business aspect and attempts to influence the behavior of people in a social context as well.

RepCheck primarily emphasizes the reputation aspect of the framework. The introduction on its home page reads “In this electronic age, where face-to-face meetings are few and far between, reputation is becoming progressively more difficult to ascertain, but no less important.”[66] The company places emphasis on the benefits of knowing about the reputation of other entities, which obviously implicates some financial incentives, especially for business entities.

Reputation is, however, the limit of the extent to which RepCheck implements the framework. RepCheck does not meet the membership and relationship factors of the framework. RepCheck’s system fails to provide a forum for open communication. It is unclear what kind of information is collected and how. It is also unclear whether other members can respond to comments made by other members, or whether RepCheck employs a numerical system similar to BizRate’s. This underrates the value of membership. The system, as established by RepCheck has a secretive, behind-the-back feel to it. A sense of cooperation is lacking and thereby the incentive to resolve issues is also lacking. In a context outside of a business one, it seems more likely that members will spread false information because there is no benefit for them to remain honest with each other. Further, if RepCheck relies on its members for information, the accuracy of such information will depend heavily on the number of members that RepCheck has, especially when one considers the broad reach over the population that the company is striving for.

Another fundamental fault with RepCheck’s system is that anonymity is missing. Perhaps members online can mask their identities, but if ratings are derived from member opinions, it would be easy for one member to narrow down the few likely candidates that may have spread the information. This is possible because most people know who is likely to talk about them because of the likely relationship that the parties already have. This is potentially a lethal flaw: members who can determine the real-space identity of the people who have rated them may be able to sue. The ability to remain anonymous therefore depends not only on the popularity of the person being reviewed (i.e. making it harder to figure out who is communicating the information), but also on ensuring that reviews are kept anonymous.

Despite the gaps in its system, RepCheck is an interesting example in that it creates a centralized location for researching the reputation of other entities, both commercial and private. The strong emphasis it is attempting to instill in members of cyber-space is perhaps a precursor to the success of reputation as a means of enforcing obligations in the future. It remains to be seen whether RepCheck will be a successful business and whether its system will become accepted by the Internet community.

IV. CONCLUSION

Our society has yet to implement strong privacy on a large scale. The problems presented when large-scale implementation is achieved requires that new means for enforcement of obligations created online with complete anonymity be created. Since the current legal system will be incapable of dealing with the problems created by strong privacy, a non-legal system must be implemented.

A non-legal system is not a new concept. It has been used successfully for many years by the cotton and diamond industries, as well as by the residents of Shasta County. The success of these systems relies on three factors: membership, relationship, and reputation.

Such a system is capable of enforcing obligations online by creating an environment where incentives exist to keep inter-member dealings honest, and where relationships are expected to last a long time. Emphasis on reputation provides a signal to members about the likelihood of success of a transaction.

Some businesses currently employ elements of the framework with success. These businesses foreshadow the likely success of a reputation-based non-legal system in the future when strong privacy is the common way of conducting business online. In my opinion, reputation-based enforcement mechanisms provide an elegant and efficient way to enforce the contractual obligations of entities that use strong encryption. Perhaps in a few years, this framework can be tested in the real world.

[1] Griswold v. Connecticut, 381 U.S. 479, 484 (1965) (finding privacy within the Third, Fourth, Fifth, and Ninth Amendments).

[2] Keith Regan. “Amazon Announces Controversial Privacy Policy.” E-Commerce Times. 1 Sept. 2000. http://www.ecommercetimes.com/perl/story/4180.html. Accessed May 3, 2002.

[5] Posner, Eric A. Law and Social Norms. (Cambridge: Harvard University Press, 2000) 108.

[7] Ellickson, Robert C. Order Without Law: How Neighbors Settle Disputes. (Cambridge, Harvard University Press, 1991) 57.

[15] Lisa Bernstein. Private Commercial Law in the Cotton Industry: Creating Cooperation Through Rules, Norms, and Institutions. University of Chicago Working Paper Series. (forthcoming) (John M. Olin Law & Economics Working Paper No. 133 (2d Series) at 3, available at http://www.law.uchicago.edu/lawecon/index.html).

[47] Bernstein Cotton 30 (Quoting telephone interview with Merchant #8).

[48] Bernstein Cotton 30 (Quoting telephone interview with Merchant #4).

[53] See “eBay and EachNet Team Up in China.” Company press release dated 3/18/02 found at http://www.shareholder.com/ebay/releases-2002.cfm.

[54] For the period ending 9/30/01, eBay reported a net profit of $18.8 million according to the company’s 10Q filing.

[55] As of 3/26/02, eBay’s stock was trading at $56.520 with a market cap of approximately $15.5 billion according to finance.yahoo.com.

[56] http://pages.ebay.com/help/community/values.html. Accessed 3/25/02.

[57] http://pages.ebay.com/help/index.html. Accessed 3/25/02.

[58] See “IBM and eBay Forge Broad Business Alliance”. Company press release dated 9/6/01 found at http://www.shareholder.com/ebay/releases-general.cfm.

[59] http://pages.ebay.com/help/community/index.html. Accessed 3/30/02.

[60] http://www.squaretrade.com/spl/jsp/eby/eb.jsp?marketplace_name=ebay&campaign=EBY_OD_6. Accessed 3/30/02.

[61] http://pages.ebay.com/help/community/insurance.html. Accessed 3/30/02.

[62] http://www.bizrate.com/content/ratings.xpml. Accessed 3/30/02.

[63] http://www.bizrate.com/content/ratings.xpml. Accessed 3/30/02.

[64] http://www.repcheck.com. Accessed 3/30/02.

[65] http://www.repcheck.com/learnmore.php. Accessed 3/30/02.

PROFESSOR DAVID FRIEDMAN

SANTA CLARA UNIVERSITY SCHOOL OF LAW

I. THE PROBLEM

Strong Privacy and Public Key Encryption

The Benefits and Costs of Public Key Encryption

The Current Legal System

The Difficulty of Enforcing Contracts Online

II. AN OLD SOLUTION TO A NEW PROBLEM

Modern Low-Tech Examples of Non-Legal Enforcement

Shasta County

The Diamond Industry

The Cotton Industry

Why Non-Legal Enforcement Works

Membership

Information Exchange

Dispute Settlement

Financial Benefits

Social Function

Relationship

Discourage Disputes

Economic Benefits

Reputation

Economic Benefits

Encourage Fair Dealing

Avoid Conflict

III. REPUTATION AND STRONG PRIVACY

The Framework

Eliminating Reliance on Real-Space Identities

Reliance on Reputation

Modern Examples

eBay

Bizrate

RepCheck

IV. CONCLUSION