Which privacy right is more important: the right to prevent your telephone number from being disclosed to others, or the right to know who is calling you? As telephone companies across the U.S. prepare to introduce a service celled Caller Identification (Caller ID), and as opponents of the service raise concerns about privacy implications, that is how the issue is being framed.

Caller ID works as follows: whenever a telephone whose owner subscribes to the service is called, it is provided with the number of the calling telephone. What the called telephone does with that number depends, of course, upon what it is designed to do with it. Most of the telephones in use now couldn't do anything with the information. Most Caller-ID compatible telephones would simply show the calling number in a small display while the telephone is ringing and during the ensuing conversation. Fancier telephones would compare the number to a stored list, either to determine the caller's name so they can display that instead of the number or to decide how to handle the call, e.g., to ignore the call, activate a telephone-recorder, play a caller-specific message, or ring the telephone. More advanced telephones, integrated into computer systems, might send the number to a database, either to add it to a list of callers or to retrieve other information about the caller.

An important point about Caller ID is that whether your telephone number is given to people you call depends upon whether they subscribe to the service, not whether you do. Subscribing lets you see the numbers of people who call you.

Caller ID has been available in private PBX telephone systems for internal calls for several years. It was made available to many businesses offering "800" numbers two years ago. Within the past year, telephone companies have begun to make Caller ID available to residential customers. By 1992, telephone companies are expected to have it available nationwide.

Proponents of Caller ID maintain that knowing who is calling is useful information that can help

prevent telephone calls from being as intrusive as they now are. They argue that if someone interrupts you by calling you, you have a right to know who they are. Furthermore, they claim, you have a right to know before you pick up the telephone, so you can decide if you want to let that person interrupt you. Also, Caller ID is touted as a way to reduce the incidence of obscene or harassing calls, since the perpetrators of such calls rely upon anonymity.

Apart from allowing telephone users to control telephone interruptions and deter crank calls,

proponents claim that Caller ID has other advantages. For example, it allows people to have telephones that can be "programmed" to handle calls from different people differently. For example, calls from relatives can be sent to the telephone recorder, calls from friends can ring through, calls from fellow bridge club members can be greeted by a message stating the place and time of the next meeting, and calls from others can be ignored entirely. Caller ID also enables "Call last caller back" for cases in which, for example, the telephone rings while you are in the shower, or is ringing when you come home with your arms full of grocery bags.

Critics of Caller ID claim that its benefits are outweighed by its violation of everyone's right to control to whom they give their telephone number. They argue that if someone wants your number, they should ask you for it rather than getting it without your consent. Caller ID opponents point to a number of legitimate types of telephone calls for which anonymity is important; e.g., people calling AIDS or child-abuse hot lines or police tip lines, social workers and medical workers calling patients from home, probation officers calling their charges, people calling businesses who don't want their number added to a telemarketing database.

Caller ID opponents also claim that the benefits of Caller ID are less than claimed anyway, because:

1. Caller ID may result in an overall increase in unwanted calls, since many businesses will use it to compile lists of customers for use in telemarketing.

2. Having Caller ID won't really reduce the intrusiveness of telephone calls. Once you've stopped what you were doing to go to the phone to see if you want to answer it, you've already been interrupted. Furthermore, it is difficult to tell from a displayed number who is calling, especially quickly enough to decide what to do before the caller gives up; people with number-display telephones will end up answering most calls anyway.

3. Most of the benefits of Caller ID can already be achieved in other ways, without revealing anyone's number without their consent. Telephone calls needn't be so intrusive. Our compulsion to answer the telephone when it rings is what should be changed: people should feel freer to get unlisted numbers and screen calls through answering machines. If the telephone company can give you the caller's number then they can trace crank and obscene calls. They can also provide "Call last caller back" without telling you the number.

Caller ID advocates rebut that the benefits of the service are real, and that those who oppose it are just "technological reactionaries" out to impede progress. Some Caller ID proponents counter critics' concerns about privacy by claiming that the service doesn't violate anyone's privacy rights, since the telephone company, not the telephone user, "owns" telephone numbers.

Caller ID critics respond that pro-Caller-ID arguments are just a smoke-screen; Caller ID's true

beneficiaries, and its main promoters, are businesses who see it as a valuable tool for collecting

telemarketing data, telephone service companies who will sell the service to businesses, and telephone equipment manufacturers who look forward to a situation in which most of the telephones in the country are suddenly obsolete.

This is an issue over which there has been, and will continue to be, heated debate. (One indication that Caller ID is a hot topic is that it was the subject of a recent discussion on "Nightline," the ABC News talk show that deals exclusively, it seems, in controversial issues. Similarly, a column on the subject by William Safire in The New York Times [December 11, 1989] drew many letters to the editor.) Telephone companies reportedly realized that some people, especially those with "unlisted" numbers, would not appreciate their numbers being disclosed by Caller ID. Nonetheless, telephone companies did not anticipate that the controversy over the service would reach the level it has. As has been said, Caller ID has been available for various types of non-residential telephone service for quite awhile. Only after telephone companies began to introduce it for residential customers did significant opposition begin to appear.

Though Caller ID has been approved for imminent availability by Public Utility Commissions (PUCs) in some areas of the U.S. and is already available in others (e.g., New Jersey), approval has rarely been uncontested. Consumer advocacy organizations, state public advocate offices, state legislators, and the American Civil Liberties Union are usually among the vocal critics of Caller ID. In areas where opposition is strong, PUCs have delayed approval or imposed complicated restrictions and conditions. In California for example, the PUC and the state assembly required that a blocking option be provided so that callers can prevent their numbers from being displayed. Pennsylvania's PUC also required a blocking option, but yielding to telephone company arguments that having such an option defeats the purpose of Caller ID, restricted its use to law enforcement officials, crime witnesses, victims of violent crimes, and others who deal with violent people. That decision has been appealed in the Pennsylvania courts by the state Consumer Advocate. Meanwhile, Senator Kohl (D-WI), has introduced a bill in Congress to require a blocking service nationwide. In the face of such complicated regulations,

some telephone companies have delayed introducing or even asking for permission to introduce Caller ID until the debate cools down.

Clearly, the idea of revealing callers' telephone numbers without their consent bothers some people, and clearly, the prospect of not having the benefits of Caller ID disturbs others. The arguments for and against Caller ID raise several issues. In what follows, several are explored in detail.

First, who really owns the rights to people's telephone numbers? Certainly, telephone companies can and do distribute people's telephone numbers in the various directories that they provide. This includes reverse directories, which customers (usually businesses) can purchase to determine peoples' names and addresses from their telephone numbers. Some directories are or soon will be available in on-line form, facilitating their use in the construction of databases.

But what about "unlisted" numbers? Doesn't the extra money a person pays for an unlisted number buy them exclusive distribution rights? Yes and no. Unlisted numbers are not included in directories, but are given out via Caller ID to whomever the holder of the unlisted number calls (provided the call recipients subscribe to the service). This argument is circular. If the telephone company can claim control of unlisted numbers because the caller ID feature makes it possible, then there's really no debate.

If you give the number of a friend to a third party without asking your friend, your friend may not like it, but you certainly would not have violated any law. A telephone company's disclosure of your number to a third party is, it might be argued, exactly analogous. However, it might also be argued that automatic or electronic disclosure of telephone numbers differs from person-to-person disclosure and is illegal. For example, in Pennsylvania, the ACLU argued that Caller ID violates that state's wiretap law, which forbids the use of devices that capture a caller's telephone number. One might also base a counter-argument not on the medium of disclosure but rather on who obtains the number. For example, one might claim that, while telephone companies can distribute directories to customers, they cannot sell lists of telephone numbers to private investigators.

So what are we to conclude about who "owns" a telephone number? Just that the issue is yet unresolved.

Next, let's examine the claimed benefits of Caller ID and the corresponding counter claims of critics. For individuals (i.e., residential telephone users), the question is: would having Caller ID enhance their privacy?

New Jersey Bell reports that the number of obscene and threatening telephone calls reported has dropped sharply since Caller ID was introduced last year. Though it is in their interest to be able to report this, it is a result that presumably could be checked. Suppose it's true. Suppose the reduction in such calls is 80%. Given that such calls represent a tiny fraction of all annoying calls, how important is it compared to an increase, say of 25%, in telephone sales solicitations, which Caller ID may well bring about?

What about Caller ID's other claimed benefits? One thing that clearly limits Caller ID's usefulness to individuals is the fact that telephone numbers identify telephones, not people. Knowing the number of the calling telephone doesn't necessarily tell you who is calling. Crank callers could simply use pay telephones to avoid detection or worse, other people's telephones, getting them into trouble. You would have no way of knowing who at a given number was calling you. If you had Caller ID, you might reject a call from your spouse stranded somewhere with a broken-down car or from someone who has discovered your injured child because you didn't recognize the number.

Another limiting factor on the overall benefit of Caller ID is the cost of the service and of the necessary equipment. Though a small percentage of the population can afford sophisticated telephones that automatically categorize numbers and handle them accordingly, most Caller-ID compatible telephones in people's homes will be of the simplest variety. Telephones or telephone attachments

that do anything more than show the calling number will be extremely rare. So, for most people, the usefulness of Caller ID will depend upon how quickly and easily they can decide, based upon a displayed number, what to do when the telephone rings. I find the arguments of critics that by the time you're looking at the number you've already been interrupted and that there will be a temptation to answer most calls "just in case" quite convincing. If they are right, Caller ID will, for average citizens, be little more than a toy that lets them answer the telephone, "Hello, Fred."

One thing proponents and critics of Caller ID agree upon are its telemarketing benefits; what they disagree upon is whether those benefits are worthwhile. According to Calvin Sims, a reporter for The New York Times, "Phone companies expect the service to be popular among businesses, which could link it to computer files of customer records." This expectation is already being realized: as stated earlier, many businesses offering "800" numbers already have Caller ID, allowing them to capture the numbers of callers. It will soon be available for companies offering "900" services. Sims says that "American Express uses an AT&T system that lets service representatives' see a caller's name and account information before they answer the telephone." With reverse-directories available from telephone companies, businesses could use captured telephone numbers to fill-out their marketing databases.

Having examined the benefits of Caller ID, let's consider some alternatives that might provide some of the same benefits without the cost to public privacy.

New Jersey Bell says that Caller ID reduces obscene and harassing calls, and calls it "the best technology" for doing so. Is it? Do you really want to know the telephone number of an obscene caller? Your telephone company can certainly get it if you report the disturbance. Shouldn't they and the police be the ones to handle such problems? Obscene and threatening calls would also be "sharply reduced" if telephone companies simply encouraged people to report them and made it clear that tracing numbers is trivial. Even if tracing numbers is not trivial now, any telephone company that can provide Caller ID could make it trivial.

Setting aside criminal calls, let's consider alternative ways to reduce "ordinary" unwanted calls. I'll begin with ways that are available now, then discuss some possible technological alternatives to Caller ID.

As stated earlier, we can simply not answer the telephone when we don't want to be disturbed. Bells can be muted; telephones can be unplugged; answering machines work as well when we are at home as when we are out. If those solutions seem dishonest or require too much behavior modification on our part, consider this: for $69.95, you can buy a device that adds a 3-digit access code to your telephone number; those who don't know it hear a perpetually-ringing telephone and you hear nothing. Arguably such a device is better than Caller ID: people you give your code to can get through regardless of what telephone they call from, people who don't have your code can't, you avoid repeated interruptions of having to decide when the telephone rings whether to answer or not, and no one's telephone number is disclosed without consent.

Now, let's turn to alternatives that don't yet exist, but could in the near future. I mentioned earlier that one limitation on the value to residential users of Caller ID is that telephone numbers identify telephones and not people. Another problem is that telephone numbers are effective, unique links to households and data records, which is why they are valuable to businesses and other potential violators of privacy. How might we fix these problems?

First, let's consider what not to do. The way the problem is stated "telephones identify telephones and not people" suggests that a solution might be to create a system in which telephone numbers identify people, not telephones. In other words, each person has a unique ID number, encoded on a card that can be inserted into telephones. In addition to being used for billing and other purposes, a caller's unique telephone-ID number is provided to whoever that person calls. People you call get more accurate information about who is calling them. This is in fact the direction that many phone pundits expect the telephone system to go. It is also precisely the wrong thing to do, because it increases the invasiveness of the technology to individual privacy, and it keeps the burden of decoding numbers on telephone-call recipients.

A much better solution is to change the system so that people are identified to callers only by name, not by a unique ID number. Names are more useful than any sort of number to residential call-recipients because they can be used directly. They are also less effective as links back to the caller because they aren't unique. The names "Fred J. Smith", or even "Caitlin D. Fitzsimmons," by themselves, would not be as useful to businesses as those peoples' telephone numbers. Two ways to supply names to call recipients come to mind: 1 ) Use the above-mentioned card, but send just the caller's name, not their account number, to the recipient; 2) Allow callers to "type" their names (or abbreviations thereof) using the key pad when they make a call. Method 1 makes giving a false name more difficult, but some people may not like even their name being given to the callee. Method 2 has the advantage that supplying one's name is optional, but lends itself to the use of false names (though that could be handled by agreement between callers and callees, e.g., nick-names).

Of course, neither of these alternatives satisfy the requirements of businesses. They aren't trying to screen calls. They want a link back to the caller. They need the telephone number, i.e., Caller ID itself. The question is: do we want them to have it?

Though the Caller ID debate has been framed as a matter of whether callers' right to keep their numbers private is more important that callees' right to know who is calling them, that is not the real issue. The issue has to do with the relative costs and benefits of the service for individuals and for businesses. I have argued that the benefits of Caller ID for residential telephone users are minimal. For them, Caller ID is a naive attempt to solve a difficult problem. Trying to decide how to handle calls based upon the number of the calling telephone is like trying to filter out commercials automatically when recording TV shows based upon the volume level (this has been attempted); you still hear a lot of commercials and you miss a lot of program material. I expect that many people will subscribe to the service, naively believing that it will help them screen calls, and then realize that it doesn't and that they are wasting their money.

While the benefits of Caller ID for individuals are doubtful, the costs are not: widespread dissemination of personal information in a society where such information is increasingly traded as a commodity. For businesses, the situation is reversed: the benefits of Caller ID are great while the only costs are money. All citizens pay the privacy cost while only businesses and people who can afford very sophisticated telephones get real benefits.

I would urge telephone companies, in their roles as public utilities, to place less weight upon the needs of businesses and more on those of their residential customers. I would urge them to seek alternatives to Caller ID that are more useful to the public and less invasive to individual privacy.

------------------------------

Jeff Johnson - September 2001

The short summary of what happened regarding Calling Number ID (CNID) since 1990 is: CPSR won the battle but lost the war.

In the late 1980s and early 1990s, state Public Utility Commissions (PUCs) across the U.S. were trying to decide whether to approve telephone company's requests to offer CNID service. After CPSR raised concerns in articles published in the CPSR Newsletter and in newspapers around the U.S., we got directly involved in PUC proceedings in several states. Among those representing CPSR in these proceedings besides me were: Erik Nilsson, Carl Page, Ronni Rosenberg, Coralee Whitcomb, and Marc Rotenberg.

The decisions of state PUCs varied, but some agreed with CPSR's analysis that CNID as proposed by the phone companies - with no ability to block one's number or with blocking available only on a per-call basis - violated caller's privacy. In California for example, the hearing officer recommended to the California PUC that CNID be disallowed. The Commission decided to allow it, but with such severe privacy safeguards that telephone companies declined to offer the service. In Oregon, CPSR's participation in the PUC hearings led to important privacy concessions by the phone company, so the service was approved. In Massachusetts, CPSR's recommendation was adopted: the service was approved with numbers being blocked by default unless the caller indicated otherwise.

Faced with strong privacy restrictions on CNID in many states, telephone companies petitioned the Federal Communications Commission (FCC) to issue rules on CNID that would apply nationwide. CPSR submitted written arguments to the FCC. Virtually ignoring public input, the FCC mandated per-call blocking as the default nationwide, overruling many state decisions. Several states appealed, with support from CPSR. The FCC denied the appeals, clearing the way for telephone companies to offer CNID with weak privacy protections nationwide. That is what we have today.

From 1991 to 1994, Jeff Johnson was Chair of CPSR. He is author of the recent book GUI Bloopers: Do's and Don'ts for Software Developers and Web Designers (Morgan Kaufmann).

"What's Inside"