For advocates of the liberal social ordera society in which the rights to life, liberty, and property are recognizedthe issue of privacy is a thorny one. Although we tend to believe, in general, that we ought to be able to maintain our own privacy, we recognize that information about us is, to the extent that it is property at all, a very peculiar kind of property.

And make no mistake: privacy is about information. As Alan Westin wrote in Privacy and Freedom, it is “the claim of individuals, groups, or institutions to determine for themselves when, how, and to what extent information about them is communicated to others.” For the liberal, the first problematic aspect of the issue is the uncertainty concerning who owns information about individuals, groups, or institutions. Suppose, for example, that Smith sees Jones in the supermarket. Presumably, Smith has a right to that information; the supermarket is, after all, a public place, and he has as much right to be there and to observe as anyone else does. But what right does Smith have to dispose of the information? May he sell it or give it away to anyone he wishes?

Of course, Jones is unlikely to claim a right of privacy about the information that he was in the supermarket at some particular time, so he is unlikely to contest Smith’s giving or selling that information to someone else. But such acceptance is not always the case: suppose Smith saw Jones buy a very expensive piece of jewelry. That information could have value to, say, another jeweler, who might annoy Jones with unsolicited advertising, or to a jewel thief, who might do more than annoy Jones.

Suppose Smith observes Jones’s conviction for stealing a piece of jewelry. After Jones serves his time in prison, he applies for a job as a jewelry salesman at Johnson’s jewelry store. Does Smith have the right to tell Johnson about Jones’s previous history as a jewel thief? Or must he keep that information private, thereby leaving Johnson at substantial risk?

Does Jones have any right to prevent Smith from selling or giving away that information? If Jones has a right to all information about himself, plainly none of that information may be transferred or traded by others. But if the information is Smith’s propertyas it would seem to be by most liberal theories, because Smith acquired it by his own effortthen evidently Smith has a right to transfer the property in any way he wishes.

The problem is complicated further by another characteristic that distinguishes information from other property: it can be transferred without diminution. If Smith tells Johnson about Jones’s history as a jewel thief, then both Smith and Johnson can use the property. In that respect, information is quite unlike physical or ordinary intangible property: if Smith gives Jones a hammer, a piece of real estate, or some shares of stock in a corporation, then Jones enjoys the benefits of owning those articles of property and Smith no longer does.

Furthermore, the cost of replicating information seems to be in constant long-term decline. Four millennia ago, information could be duplicated from one person to another only if someone orally related it and someone else processed it into his memory. Such transfer was both cumbersome and relatively unreliable. Two millennia ago, information could be encoded and stored, but the process (handwriting) was very labor-intensive. Half a millennium ago, information could be mass produced by printing, a huge improvement but still an expensive process unless a large number of copies were needed. A half-century ago, information could be reproduced by copying machines. For the first time, small quantities of information could be stored and reproduced in small quantities at a relatively low cost: a sheet of paper could hold approximately 5,000 characters, or 5 kilobytes of information, at a cost of about 5 cents per kilobyte. Today, anyone with a microcomputer can duplicate 650 megabytes of informationthe entire Encyclopedia Britannica, for exampleat a cost of perhaps $1.00, or about 0.00015 cents per kilobyte.

The upshot is that if Smith can get useful information about Jones, he can reproduce and distribute it very cheaply. Consequently, Smith can afford to go to considerable expense to get useful information about Jones. But he probably doesn’t have to: the same technology that makes reproduction and distribution of information cheaper also lowers the cost of its acquisition.

If Jones shops at an up-to-date supermarket, its computer keeps a record of all his purchases, and anyone with access to that information can learn a great deal about Jones: that he prefers health food, or smokes cigarettes, or drinks wine, or eats fatty foods, or buys condoms or pornographic magazines, etc. If that information belongs to the supermarket and not to Jones, as liberal theory seems to suggest, what is to keep the supermarket from selling it? What prevents marketers, or the government, from compiling an elaborate database on the entire population, complete with information about people’s medical treatment, what they purchase, where they travel, the Internet sites they visit, the content of their electronic mail and telephone conversations, and nearly every other aspect of their individual lives?

And don’t think such a database cannot be compiled. The technology already exists, and current databases maintain a substantial amount of the requisite information in very accessible files. And the data will continue to grow and to become more accessible and transferable at ever lower cost.

What is the state to do, if anything, about this situation? In Privacy in the Information Age, Fred H. Cate surveys the problem and summarizes the approaches to privacy embodied in American and Western European law. In general, the European approach is to create a powerful regulatory body to control the purchase and sale of information, whereas the American approach is, within certain boundaries, to allow the market to deal with the problem. He concludes with an intelligent speculation about the future of the law of privacy. “The government should play a circumscribed role,” he writes, “limited primarily to articulating principles for information privacy, enacting laws when necessary to protect the rights outlined above, adjudicating disputes concerning such rights, facilitating discussion, education, and cooperation, and leading multinational negotiations to resolve conflicts among competing national privacy laws.”

He adds three appendixes: the European Union’s privacy directive of October 24, 1995, a report on the National Information Infrastructure, and a summary of the discussion of his book by experts at a conference prior to its publication. The first two of the appendices contain useful information, but are a bit dry. I suspect many readers will skip them. But I am glad they were included, because they contain useful information about what is happening now in the area of privacy law.

Cate is an expert in the best sense of the word. He knows the technology and the law; he understands how information is acquired and traded in the private sector; and he has given the whole matter a great deal of thought. He judiciously keeps his conclusions tentative, but readers respectful of the liberal social order, and thus dubious about the merits of government intervention beyond the enforcement of contracts, will find a great deal of merit in his work. So, for that matter, will those who are dubious about the ability of individuals acting in the marketplace to deal with problems of this sort. Whether one agrees with Cate’s tentative conclusions or not, his mastery of the subject and the clarity of his exposition make Privacy in the Information Age a valuable and thought-provoking book for anyone concerned about either his own privacy or the way personal and institutional information is controlled as its accumulation and dissemination become easier and cheaper.