11-Jan-84 22:49:51-PST,24630;000000000000 Return-path: Mail-from: DECNET site ECLA rcvd at 11-Jan-84 2248-PST Date: 11 Jan 1984 1353-PST From: Rob-Kling Subject: Review- Rise of the Computer State To: telecom@USC-ECLC Received: from UCI-20b by UCI-750a; 11 Jan 84 14:08:56 PST (Wed) Via: UCI; 11 Jan 84 21:21-PDT Via: Usc-Cse; 11 Jan 84 22:41:56 Rise of the Computer State by David Burnham Published by Random House, New York - 1983. Review by Rob Kling Department of Information and Computer Science University of California, Irvine (KLING.UCI@RAND-relay) This book examines the ways that Federal agencies and elected officials have employed computer-based information systems (CBIS) to increase their power unfairly. Burnham's main theses are: 1) that CBIS have often been effective media for extending the surveillance potential of the host organization; 2) overall, citizens have lost substantial power in their routine dealings with computer-using organizations; 3) attempts to regulate the use of CBIS containing personal records have been frail and largely ineffective relative to the scale of operations that should be regulated; 4) some organizations which employ sophisticated CBIS for intelligence, such as the National Security Agency, are unaccountable to the larger public. These theses have a sinister tinge. As we enter 1984, the United States is far from a police state. However, Burnham fears that the slow, steady, consistent adoption of new surveillance systems and the expansion of existing ones is eroding democratic political processes in the United States. If he is correct, these are arguably the most important consequences of computerization in the United States. This is a trade book aimed at the same audience that reads "Megatrends", "The Third Wave", or "Fifth Generation". Unlike these highly popular books which are permeated with happy talk about the social possibilities of widespread computerization, "The Rise of the Computer State" examines the seamy underside of organizations that employ CBIS to collect, manipulate, and communicate sensitive data about all of us. Burnham, a New York Times reporter, has written this book for a popular audience. Its strengths lie in Burnham's sensitivity to the civil liberties issues in practices that might simply appear "expeditious" and in his eye for graphic detail in explaining how organizations employ CBIS to make their operations efficient and "effective." Burnham examines two themes that link computerization with a certain kind of organizational power: surveillance of "targeted" people or groups and opinion polling. In a separate chapter he examines the National Security Agency which he labels "the ultimate computer bureaucracy." SURVEILLANCE Some organizations act under legislative mandates that many people would label "pro-social". For example, the Bureau of Child Support of the Los Angeles District Attorney's office uses CBIS to seize California State tax refunds from certain runaway fathers who are delinquent in their child support payments. A second group act within the boundaries of legal, but unduly permissive information practices. For example, a company called U.D. Registry provides landlords with histories of disputes with previous landlords, maintains records which are usually unknown to tenants and does little to insure that they are treated fairly. A third group of organizations engage in action that are either illegal or nearly so. For example, U.S. Army's surveillance of liberal and leftist activists in the late 1960's, extended well beyond the scope of "national security." Burnham portrays these activities with sharp detail that give color to routine practices and their participants. Burnham is a staunch civil libertarian and sees all social surveillance as problematic. It is easiest to criticize organizations like the U.S. Army when they intrude upon political minorities and thereby threaten First amendment rights. It is also easy to criticize some of the "holes" in CBIS such as those operated by U.D. Registry, which are unknown to people on whom records are kept, and who are not legally obligated to enable people to see their records, correct errors, or annotate their files case of disputes. Burnham's criticisms reach much further than identifying the problems with CBIS employed by the second and third groups of organizations. He questions the first group as well. Burnham's questions about organizations and systems for tracking runaway fathers who leave their children on welfare illustrates of his concerns about social strategies which depend upon extensive surveillance for enforcement: 1) will the original target group be slowly enlarged until it is much larger than originally intended in the enabling legislation? 2) can the information system be extended by local officials for surveillance upon "others who fall into disfavor?" Burnham reports how the scope of these systems has expanded from locating parents who were avoiding child support payments and whose children were receiving funds from Federal welfare programs to include any parent whose (ex)spouse seeks the other parent of their children. Burnham notes that there are few constitutional limits on the scope of such an surveillance system. Why not, for example, expand its scope so that creditors can track down their debtors? Or why not expand it expand it so that people can locate lost relatives and old friends? While these "information needs" are less heart wrenching than the situations of women who turn to public assistance when their ex-husbands refuse to pay court-mandated child support, they are also "pro-social." Burnham argues that little prevents surveillance systems such as this one from being slowly expanded to track ever larger groups of people than legislative sentiment and a fragile coalition of legislators who are sympathetic to civil liberties values. Burnham uses this example to illustrate another key feature of recent surveillance systems: records systems which are set up for rather narrow purposes of one organization are used by investigators in another organization. The Parent Locator System, for example, is not a particular, specialized CBIS. Rather, it is a set of procedures and arrangements which enable certain investigators to send lists of "missing parents" to the Internal Revenue Service, the Social Security Administration, the Defense Department, the Veterans Administration, and the National Personnel Center. Each of these organizations honors these requests, searches its CBIS for the current locations of the "missing parents" and returns the information to the requestors on magnetic tape. While many CBIS could be operated as manual systems, these searches would be prohibitively expensive add-ons with manual record systems. However, the marginal costs of search are affordable with computerized record systems. The Parent Locator "System" is one of many "matching programs" in which public agencies use existing files to search for deviants. Organizational payroll files have been "matched" against welfare files to find gainfully employed people who are committing welfare fraud. State Department of Motor Vehicle files have been matched with Selective Service files to identify eligible 18 year olds who have not registered for the draft. In each of these cases, the records of thousands of people who have broken no laws are matched to find the few that have. Burnham finds the principle offensive, even though the applications are expedient and have so far have been aimed at lawbreakers. In his eyes, expediency and efficiency should not be preeminent values for administrative action. PRIVACY REGULATIONS Burnham briefly examines some of the Federal privacy initiatives of the last decade, including the Privacy Act of 1974, the proposals of the Privacy Protection Study Commission, and the 1978 Financial Right to Privacy Act. These laws have provided minimal protections, and important protections of the Federal Right to Privacy Act have been undermined in implementation by Federal agencies under Ford, Carter, and Reagan. Only a few of the 155 recommendations reported by the Privacy Protection Study Commission in 1977 have been enacted in law. Burnham mentions these laws and examines some of their limitations. However, he doesn't evaluate their potential. Would many of the problems of CBIS operated by firms like the U.D. Registry be ameliorated if they were brought under laws like the Fair Credit Reporting Act? Would civil liberties be better protected if the remaining recommendations of the Privacy Protection Study Commission were enacted in law? Unfortunately, Burnham is mute about these possibilities. Burnham is strongest in identifying concrete problems. Most serious there is no permanent institutional counterweight to Federal agencies when they propose new, more efficient, or enlarged personal record systems. Agencies such as the FBI, the IRS, or the Social Security Administration can return to Congress every few years with proposals for massive CBIS which have problematic privacy aspects and expect that sooner or later, the civil libertarians who restricted their last proposal will be weaker or pre-occupied with other matters. POLLING Burnham examines opinion polling as another form of organizational intelligence which has been rendered substantially more efficient and sophisticated by computers. He views opinion polling by elected officials and organizations which are campaigning for specific legislation as selective intelligence which places the target public at an unfair disadvantage. The main problem he sees in market research in the service of electoral politics is the extent to which it helps make propaganda less transparent and the public more manipulable by marketing strategists who target different messages to different groups. While there is nothing new in political actors tailoring their appeals to different audiences, Burnham fears that the modern versions of sophistry are less obvious and consequently far more successful for those who can afford to employ them. He also views opinion polls as easily subject to manipulation by politicians seeking legitimacy or publicity. Polling is not simply a reporting device. Pollers gain leverage relative to the larger public since much of the audience for polls will read headlines and short news items which distort the scientific meaning of a poll by neglecting to explain the nature of the sample, the detailed distribution of responses, or the questions asked. Political polling is not only "information gathering;" it can be a devise for persuading larger publics about the popularity of one's position. NATIONAL SECURITY AGENCY In a dramatic chapter, Burnham reports how the National Security Agency (NSA) has operated under a charter which has remained secret it was initiated by President Truman in 1952. The NSA specializes in electronic surveillance. A large fraction of its efforts probably go to observing military force deployments and strategic resources worldwide. Burnham reports how the NSA has also illegally eavesdropped on a significant fraction of international telephone calls and telex messages which leave the United States. Burnham reports on the character of specific programs of domestic surveillance which were illegal. According to Burnham, the NSA developed files on political dissidents including civil rights activists, antiwar activists, members of Congress, and ordinary citizens who were critical of official government policies. While most of the domestic political surveillance appeared to take place in the late 1960's through mid-1970's, the shroud of secrecy that surrounds the NSA makes it difficult to have significant Congressional oversight of its policies and practices. During the last 5 years, the NSA has moved to control cryptographic research in the United States. Recently developed encryption schemes are based on sophisticated algorithms which require digital circuits for rapid coding and decoding. Some of the new schemes even allow the code keys to be public, rather than secret. As more business operations in the United States is computerized, organizations seek ways to protect the privacy of data such as large funds transfers. Thus the market for efficient and effective data encryption devices has expanded beyond the intelligence community to include financial institutions. The NSA has recently taken control of this research out of the hands of the National Science Foundation, even though it has no publicly documented legal mandate for its action. There are deep policy questions about whether national security is well served by the availability of cheap encrypting devices which are effectively unbreakable. These questions are not being raised in public debates, nor does Burnham shed much light on them. Rather he simply adopts the libertarian critique of surveillance. Like other political labels with strong moral content, it has been abused as a cover for unsavory actions carried out by government operatives. The term in not wholly vacuous and Burnham glosses most of the knotty policy issues. TOWARD A POLICE STATE? Burnham's theses are loosely fabricated from dramatic examples. He does not offer explicit hypotheses, strong organizing concepts, and a way of placing his examples in a context which enables a reader to understand their overall significance. Examples of bad outcomes can elicit sympathy for "victims." But systematic information about the frequency and extent of problems and abuses are necessary to demonstrate that the overall social setup within which they happen is badly flawed, corrupt, or perverse. Some of his examples of people victimized by slips in CBIS and organizational practices suggest that Kafka has provided better guiding images than Orwell for appreciating a computer-based, mobile, organizational society. Burnham has little taste for irony, and explores Orwellian abuse rather than Kafkaesque happenings. Do the events Burnham describes indicate that Federal agencies and other large computer users are pushing the the US along a path of political development that is leading to a much less democratic form of Federal government? Unfortunately, Burnham does not describe the changing nature of Congressional oversight and public accountability sufficiently well to provide a clear answer to that question. He succeeds in generating sentiments in favor of this hypothesis by his accumulated cases of organizational seaminess and occasional abuse. But he relies heavily upon a reader's distrust of elected officials and large bureaucracies to help cement his case. He also relies upon general theses about power, such as Lord Acton's maxim. Unchecked power often corrupts and organizations are often less willing to be fair to their clients than efficient and autonomous. But general principles do not make the particular case since the variations in actual organizational practices are significant and vast. One peculiar feature of contemporary police states, such as those in Eastern Europe and Latin America, is the extent to which they have relied upon low technologies for extensive social control and even mass terror. Many abusive ruling cliques rely upon neighborhood informants, secret trials, and mysterious disappearances to maintain control. They don't need database management systems, teleprocessing, and spy satellites. Low technology strategies are especially effective in "small town" societies. Burnham's implicit argument is that less obtrusive forms of surveillance and social control can harm the political culture of liberal democracies. CBIS are attractive to administrators and politicians because they promise heightened efficiencies and sometimes enhanced fairness in providing services to large mobile populations. However, the anecdotes of errors with a human cost and even abuses which Burnham piles on the reader, illustrate problems but do not make his case. Burnham's strongest case is his critique of the NSA's abuses of authority. Like, the secret Law Enforcement Intelligence Units, much of the problem with the NSA comes from its shroud of secrecy and freedom from significant legislative oversight. It's use of computer-based monitoring systems is incidental to its problematic place in American political life. I suspect that one basic issue is accountability of these agencies to the public through the legislatures. At times this is no easy task when the administrative agencies can shroud their actions with the complexities of high technologies. There is a strong case to be made that in the clashes between branches of government, administrative agencies have found legal and technological loopholes to temporarily free themselves from regulatory restraint. Congressional actions are not always right. But there is an argument that administrative agencies have been able to exploit computer-based technologies to shift the balance of governmental power away from elected officials. This systematic shift of power has been best documented in the case of local governments. It is likely to be happening at other governmental levels as well. COMPUTERS AND POLITICS Burnham is sensitive to the shifts of power to executive agencies. But he is at a loss to explain them very well. He misses the deeper politics of computing. I find a clue to his misperception, a very common one, in his reference to "the computer's system of thinking." For Burnham, CBIS are simply highly structured, logical, possibly hierarchical information processing "tools." He misses the ways in which CBIS designs often reflect the "systems of thinking" of those who propose them. CBIS promoters may label their preferences as "required by computers" to help their case, but they often ignore or discourage many technical and administrative alternatives. Many CBIS are usefully viewed as forms of social organization. They are composed of many layers of data, programs, and communications support stretched across different organizations. Those who oversee them need some ability to appreciate technical alternatives and also have some substantive expertise in the organizational functions which have computer support. This dual expertise is rare, particularly among elected officials at all levels of government. As a consequence, they have trouble in providing sensible guidance to executive agency staff. QUALITY OF BURNHAM'S ANALYSIS I would like to like this book more than I do. I like Burnham's eye for detail and his relentless questions about the underside of computer-based surveillance systems. Some new data brokering organizations start up each year. Each year, many existing organizations expand the scope and scale of their record keeping. Laws and administrative practices also change slowly each year. Over ten year periods, these gradual small scale changes accumulate. Periodic reviews of these practices are useful. As a consequence of continuing changes in organizational practices, legal arrangements, and technology, studies published in the early 1970's such as Westin and Baker's "Databanks in a Free Society" or James Rules' "Public Surveillance and Private Lives" have become dated. Both of these studies pre-date the use of computer matching, and several Federal privacy initiatives. Unfortunately, this book is weak in analysis. Even the chapter headings don't guide the argument. The first three chapters are labelled "surveillance," "data bases," and "power." However, themes of power, surveillance, and data bases are strong elements in each of them. The chapter labelled "power" primarily examines political polling. This lax labelling of chapters signifies the way that Burnham eschews tough analysis in favor of easy sentimentalizing. It should be hard for Burnham, a reporter and hence a kind of intelligence agent, to find observation, reporting, and persuasion to be inherently sinister acts. However, Burnham colors his narrative so that people who administer a CBIS are stigmatized in descriptions such as "(speaking) in the quiet monotones of many long-time government employees," or are "slightly Mephistopholean." People who sympathize with civil libertarian values are portrayed without any frailties. Burnham is deeply suspicious of pollsters and politicians who manipulate the public with numbers, but he is very adept at manipulating his audience with images. These images which equate personal goodness with political philosophy grossly mislead. Despite these limitations, "The Rise of the Computer State" is particularly important because it helps articulate and illustrate important questions about computing and social power. Unfortunately, there is no other up-to-date inquiry into organizational surveillance and high technology. "The Rise of the Computer State" is an important contribution to the tiny stream of literature which examines the political dimensions of computer-based technologies in public life. I hope that many people read this book despite its analytical flaws. It helps dramatize the current problems of computer-based surveillance systems. Burnham's graphic episodes can help give people who are not intimate with CBIS a keen appreciation of the dilemmas which their use and users are now creating. It can help more systematic investigators by providing a rich set of clues from which to continue this inquiry. These questions will not go away after 1984 has passed.