In early 1963, the IRS installed a new IBM model #7074 computer inside a “starkly antiseptic” 40-foot room just east of Martinsburg in the West Virginia hills. The machine was a rental, leased at a hefty $4,000 per day, but IRS Chief Mortimer Caplin was confident the increased efficiency it brought to tax collection would far outweigh the cost. When news of the upgrade went public, cartoons in the press depicted bleak Orwellian scenes of taxpayers being spied upon by mechanical brains and gigantic machines with long tentacles weaving in and out of living rooms, banks, and the U.S. Treasury. Popular Science reported that “as far as the IRS is concerned, your personality, history, and spirit . . . will be reduced to magnetic specks on a single reel about the diameter of an LP record,” and predicted that soon large corporations might also plug into “Big Brother 7074.” Representative James Oliver (D-Maine) eloquently expressed this same sentiment years earlier in a House Subcommittee on Government Statistics: “It’s my impression that these machines may know too damn much.”^{Footnote 1}

Contemporary observations about privacy and data collection in the United States, now one of the foremost concerns facing our society, are legion. We decry the privacy problems raised by the digital age while simultaneously wondering what solutions, if any, are at our disposal to balance these concerns with the clear advantages in efficiency and convenience digitization affords. Yet as this conversation matures in the academic community, historians have remained noticeably silent about information privacy. They have left the heavy lifting to legal scholars, journalists, political scientists, and technology experts. This lack of historical engagement is troubling when we consider that the current state of data collection was largely a product of political, social, and cultural developments that date back to the Eisenhower era. Americans have been debating this issue for more than half a century.^{Footnote 2}

Privacy is, by nature, a multifaceted concept, and the current literature on information privacy is far from monolithic. Modern tech studies have been very useful in chronicling the extraordinary growth of information systems and the increasing sophistication of data processing. They bring much needed attention to the impact of digitization on personal privacy and better equip us to understand the capabilities data holders have to use information in ways we may find objectionable. But these works are almost always rooted in presentism and can mischaracterize the vast expansion of data collection by public and private entities in the United States as no more than an inevitable byproduct of technological evolution.^{Footnote 3}

The legal community has given considerable attention to issues of privacy over the last twenty years, so much so that information privacy has been legitimized as a distinct field of legal inquiry. The search for legal remedies to modern privacy problems has yielded a better understanding of where the right to privacy is located in our liberal democratic framework, where the law has not kept pace with technology, suggestions for appropriate enforcement mechanisms, and a more precise conceptualization of what privacy means. These contributions have been invaluable in many regards, but they also tend to ignore how perceptions of privacy are shaped by social and cultural forces, often get bogged down in debates over definition and precision, and can overlook the fact that legal proscriptions are just one among a variety of factors that influence the decisions of data holders.^{Footnote 4}

In the realm of political science, scholars have sounded the call for a more rigorous engagement with information privacy by positioning it as a concept rooted in political theory and public administration. Among the more notable contributions have been profiles of the various actors and strategies employed by modern privacy advocates and broader comparative studies of international privacy policies that seek a better understanding of why, as Colin Bennett and Charles Raab have pointed out, “the United States is now the only advanced industrial state that has not passed, or is not in the process of passing, a [comprehensive] data protection law covering private sector activities.” But even these scholars are quick to note that the analysis of privacy and public policy tends to be ahistorical, that “the historical perspective on today’s and tomorrow’s policy issues risks being lost,” and that the field would be better “enriched” by historical analysis.^{Footnote 5}

Historians can help fill these gaps by engaging the longer narrative of information privacy in the United States and placing particular emphasis on how the collection and processing of information flourished within distinct patterns, conflicts, and contexts that were subject to historical forces. Such work would contribute a better understanding of where Americans have already been on this issue, what has changed or is in the process of changing, and what has stayed the same. History shows us that as the United States stood on the cusp of the Internet age, its privacy advocates failed to articulate an idea of privacy capable of formulating effective policies and protections. It also shows us that privacy, despite these weaknesses, made great strides as a policy issue for a variety of reasons. The how and why are important.

While a thorough survey of data collection in the United States is too exhaustive to tackle here, three key points of entry and a familiarity with the basic narrative might allow historians to better engage this topic.

First, we must recognize that two developments—the advent of mainframe computers in the 1950s (known affectionately as Big Iron) and the rise of the American welfare state—are most responsible for elevating the privacy problems of public data collection to a position of national significance. Taken together, their intersection marks a critical benchmark in the history of American privacy.

Governments must first acquire information about their citizens before they can effectively extend them social and legal benefits. As the bureaucratic structure of the United States expanded throughout the twentieth century, the amount of information collected about Americans grew exponentially. Big Iron made it possible to streamline this data. It provided the means through which the information held by each department could be stored, accessed, and shared more easily, and the means through which this data was later shared with commercial entities on a massive scale.

But while the aspirations of New Deal liberalism played the greatest role in the unprecedented expansion of data collection in the United States, it was, ironically, attempts by fiscal conservatives in the 1970s and 1980s to reign in “big government” and the “nanny state” that brought a marked increase in the aggregation, processing, and sharing of that data to a point where the number of accessible individual state and federal files reached into the billions. The conservative ascendency, particularly the Reagan administration, expanded the practice of information collection and computer matching in the name of efficiency, free enterprise, and “small government.” Powerful private interests who benefited from access to government data aided the effort.^{Footnote 6}

Second, historians must understand that access to sophisticated data collection technologies grew increasingly democratized over the second half of the twentieth century. The earliest computers were so expensive and cumbersome to maintain that only large organizations with vast funds at their disposal such as government agencies, banks, and telecommunications companies could maintain large-scale databases. As computers became more affordable, a host of smaller actors were also able to participate in data collection, which in turn put greater pressure on individual privacy. This democratization of data inhibited the ability of privacy advocates to enact adequate protections against the improper circulation and misuse of personal information. Instead, they were forced to settle for “sectoral” or patchwork policy remedies.

Privacy advocates in the 1960s gained a measure of success largely because the threat to privacy was the familiar threat of those in positions of great power. As time proceeded and databases become more democratized, they came to face not just the government but a host of commercial entities with varying aims and practices that made the privacy problems they faced more confusing, harder to articulate, and more difficult to address with policy initiatives. Information collection became more mobile and profitable, and the forces privacy advocates opposed were less grounded in familiar social and political traditions. They also conflicted with free-market ideologies. This is still the case today. As privacy expert Helen Nissenbaum argues, “This confluence of technology with a particular power configuration has broken down with the democratization of access to automated record keeping and related technologies.”^{Footnote 7}

Third, the privacy protections enacted by Congress throughout the twentieth century worked more to enhance the extension of data collection than they did to constrain it. Privacy expert James Rule notes that making data collection systems more open and more accountable in the interest of balance and fairness tends to “disarm public objections and streamlines acceptance of the underlying practices” and “by establishing that any ‘purpose’ or ‘need’ for personal information justifies such monitoring,” privacy codes often “open the floodgates for unlimited extension [of these practices].” Privacy protections achieved through legislation, in the opinion of some legal scholars, have largely served to “smooth the bumps and brush aside obstacles en route to a vastly less private world.”^{Footnote 8}

Before beginning, historians also require some clarification on problems of definition. Privacy is a word. Generations of scholars have failed to produce an interpretation of that word that is both concise and definitive. Recent scholarship in legal theory outlines a more sophisticated taxonomy of privacy that abandons the search for a common denominator in favor of something more akin to Ludwig Wittgenstein’s notion of “family resemblances.” Daniel Solove suggests that “certain concepts might not have a single common characteristic; rather, they draw from a common pool of similar elements. Privacy, therefore, consists of many different yet related things.” Nissenbaum, as well, urges scholars to view informational privacy through a framework of “contextual integrity,” and consider that “informational norms evolve over time in distinct patterns from society to society,” that “what people care about most is not simply restricting the flow of information but ensuring that it flows appropriately.”^{Footnote 9}

As Americans debate the multitude of privacy problems in the digital age, it is essential that we first have some semblance of how we’ve conceived of privacy and data collection before its arrival. A broader chronology is useful to this end. By not exclusively focusing on specific thematic incidences (such as the ’74 Privacy Act or 1980s data mining) we can better identify the pervasive nature of information privacy as a theme in American political culture.

IBM’s meteoric rise in the tech market was largely due to the federal government’s increasing need for data processing as it expanded the American welfare state. About the time the New Deal was creating a slew of new federal agencies, the United States became one of the company’s best customers. The storage and processing needs of the postwar American government made it a pioneer in the development and use of electronic data processing. By 1966, twenty-two federal agencies had installed mainframe computers the size of city apartments to store and access information concerning the American public and approximately 1.5 million federal employees.^{Footnote 10}

A survey of scientific and popular literature reveals little in the way of public apprehension over the effects of computers on personal lives in the 1940s and 1950s. The tipping point came in the early 1960s. As the collection and storage of data grew more sophisticated, voices in the technology community began to voice concerns over a lack of privacy safeguards. Two books published in 1964, Vance Packard’s The Naked Society and Myron Brenton’s The Privacy Invaders, were among the earliest attempts to call attention to the unchecked expansion of computerization. Packard argued that “the expectation that one has a right to be let alone—the whole idea that privacy is a right worth cherishing—seems to be evaporating among large segments of our population.”^{Footnote 11}

Despite this blowback, there were still many in the academic and tech communities who favored computerization and began to complain the process of government recordkeeping wasn’t as efficient as it might have been. Data were scattered across dozens of agencies, collected on an inconsistent basis, and were largely inaccessible to any single group of researchers. Because cross-referencing data took weeks if not months, statisticians felt the lack of centralization was an unnecessary obstacle that stifled Big Iron’s true potential.

In 1965, a three-year study completed by the Social Science Research Council (SSRC) of the American Economic Association argued that “neither outside scholars nor federal agencies [were] able to utilize public data efficiently because of excessive decentralization in the maintenance of the data files.” What came to be known as the Kaysen Task Force Report strongly advocated the creation of a centralized national data center to be drawn from the records of the Census Bureau, the Bureau of Labor Statistics, the Internal Revenue Service, and the Social Security Administration. Opposition to the proposal was fierce and immediate, and a special subcommittee of Congress was tasked with investigating the privacy problems raised by the creation of such a center.^{Footnote 12}

The first House subcommittee hearing dealing with the computer and invasions of privacy convened in 1965. “Today the computer is a central figure of our society,” opined Representative Frank Horton (R-N.Y.), “the increasing rate at which it will change our lives exceeds the imagination, exceeds even the imagination of the computermen who foster it.”^{Footnote 13}

Advocates for the data center contended the proposal was merely the next logical step by a government that already collected, organized, and published vast amounts of data in response to both state and public demands for useful information. That to argue against the center was to argue against the inevitable. Others positioned themselves as champions of the public good, noting that it was “becoming increasingly difficult to make informed and intelligent policy decisions on such questions in the area of poverty as welfare payments, family allowances, and the like, simply because we lack sufficient ‘dis-aggregated’ information.”^{Footnote 14}

But the data center’s supporters were outgunned. Opponents raised a number of effective objections under the mantle of privacy—objections that drew heavily from well-established symbols in popular literature and science fiction. Central to the arguments of many privacy advocates was the image of the “Computerized Man,” most famously articulated in George Orwell’s 1984. The metaphor was so powerful and so readily apparent that subcommittee chair Congressman Cornelius Gallagher (D-N.J.) used it to open the proceedings, describing a protagonist “stripped of his individuality and his privacy. . . . His life, his talent, and his earning capacity would be reduced to a tape with very few alternatives available.” Another witness claimed that “Big Brother, if he ever comes to these United States, may turn out not to be a greedy power seeker, but rather a relentless bureaucrat obsessed with efficiency.”^{Footnote 15}

The 1984 image was engaging, but the more substantive arguments from privacy advocates moved away from the realm of fiction toward the permanent nature of the data collected, the lack of context associated with a computer printout, and the dangers of an inescapable past.

Information gathered in small scraps and later reassembled into a single file generally did not include the context in which each piece was collected. This practice often prevented the subject from understanding the larger implications of the information they’re sharing, and illustrates a distinct privacy problem Solove categorizes as “information aggregation.” “Human beings,” argued Paul Baran of the Rand Corporation, “by their day-to-day need to make decisions using totally inadequate evidence are prone to jump to conclusions when presented with very thin chains of inferred relationships.” Frank Horton thought “pigeonhole” was a more fitting verb to describe this phenomenon.^{Footnote 16}

The data center’s opponents were not Luddites. They knew it was foolish to ignore the natural trend toward centralization and they knew that technology only ever follows one trajectory. But they also knew that until very recently, privacy’s strongest safeguard had been the fragmented nature of the information held by private and public agencies. The specific privacy problem at hand was not one of new information being disclosed, but one rooted in what Solove’s taxonomy calls “increased accessibility,” or the wider dissemination of information that was already available.

Perhaps the most influential voice in the privacy debates of the 1960s was Alan F. Westin, author of two core works on the subject that paid special attention to the discussion of data collection. Westin defined privacy as “the claim of individuals, groups, or institutions to determine for themselves when, how, and to what extent information about them is communicated to others.” He advocated moderation and pragmatism, noting that while privacy was important, the Orwellian imagery used by privacy advocates was largely inappropriate as most who supported computerization were not tyrants but well-meaning actors in an age of increased social consciousness who were caught up in an “exciting challenge” to “know more about the factual universe than had ever been possible before.” He noted that data collection clearly required some kind of regulation in the face of technological development, but that the role of policymakers must be to provide a balance between privacy and computerization. That as a society “we must support both science and liberty” through proper legislative safeguards.^{Footnote 17}

Privacy advocates won the data center debate. The Eighty-Ninth Congress rejected the motion. But the absence of a national data center did little to hinder the larger forces privacy advocates were pushing against. The practice of collecting and sharing personal information grew exponentially in the late 1960s and early 1970s within both the public and private sectors. Credit bureaus played an increasingly important role in the financial lives of individuals. American corporations, as well, used data gleaned from government agencies and other companies to make their direct-marketing campaigns more sophisticated. Many insurance companies and investigative employment firms actively pursued every bit of data they could get their hands on to make informed assessments. And government agencies continued to lean on Big Iron as a way to streamline their records.

At its core, data collection is about discrimination, “discrimination in determining precisely what actions are warranted to each member among large populations” according to privacy expert James Rule. With proper data, organizations can discriminate as to “who is worthy of credit, and how much credit should be extended” or discriminate between “individuals more or less dangerous to the public order,” or as to “who is liable for tax assessments or social insurance benefits, and to what extent.” In the private sector, the reward for effective discrimination is profit.^{Footnote 18}

With the profit motive firmly in place, data processing expanded from the late 1960s into the 1970s through the increased sharing of aggregated data between interested commercial entities and government agencies. Because it is considerably cheaper to share information than it is to reproduce it, public and private organizations facilitated symbiotic relationships across multiple information sectors. Bank account information is of interest to credit card companies. Medical information is of interest to insurance companies. Tax returns are of interest to mortgage lenders.

Responding to growing public concern (and the scholarship of Allen Westin), Congress passed the Fair Credit Reporting Act in 1970, the first information privacy legislation in the United States. The law required credit agencies to disclose their information to consumers looking to check their personal report for errors, and established penalties for the unauthorized disclosure of private information. Many in the media and other public figures balked at the law’s ineffectiveness. Sheldon Feldman of the Federal Trade Commission (the agency tasked with enforcing the law) remarked soon after its passage that the legislation was rather “fuzzy” on explaining why an individual might be denied credit.^{Footnote 19}

Despite the Fair Credit Reporting Act, insurance companies routinely checked with their customers’ neighbors about perceived drinking habits, home life, and leisure activities. Private investigators and employment firms checked the arrest records, tax history, and car payments of their client’s prospective applicants. Credit bureaus sought out every financial transaction made by their subjects. And sales divisions poured over data purchased by these organizations to better identify prospective buyers and so tailor their direct-marketing campaigns.

By the early 1970s books like Charles Fried’s An Anatomy of Values (1970) and Arthur R. Miller’s The Assault on Privacy (1971) contributed to a growing body of journalistic and academic literature on information privacy that looked to educate the public about the vast gains made by public and private databases in recent decades. Between the data center proposal and 1974, there were twelve hearings in the Senate, fifteen in the House of Representatives, and two joint hearings in Congress on privacy and data collection. One 1971 Senate subcommittee identified 858 databanks used by 54 federal agencies operating with disturbingly unclear oversight.^{Footnote 20}

In 1973, Miller helped author a substantial report by the Department of Health, Education, and Welfare (HEW) that proclaimed, in no uncertain terms, that “under current law, a person’s privacy is poorly protected against arbitrary or abusive record-keeping” and that there existed an urgent “need to establish standards of record-keeping practice appropriate to the computer age.”^{Footnote 21}

The overall goal of the report was to explain how best to provide individuals with more control over the information that was being collected, processed, and disseminated about them. This could be achieved, it argued, through the implementation of “fair information principles,” a series of information practices Colin Bennett traces to policy analysts in Europe and the United States in the late 1960s and early 1970s. Together, the five principles outlined in the report’s “Code of Fair Information Practice” hoped to balance privacy concerns and databanks through appropriate regulation. They included the prohibition of secret data-based record-keeping systems, a clear process through which individuals could access their own records and correct errors, a means for individuals to prevent “secondary use” or information obtained for one purpose being used for other reasons without the subject’s consent, and a requirement that any organization engaging in information collection take appropriate measures to prevent the misuse of the data they hold. Upon outlining their recommendations, the committee’s paramount concern was repeated throughout the report: the law was simply not keeping pace with technology and that legislation would be required to keep the practice of record-keeping from boiling over.^{Footnote 22}

The HEW report expressed other concerns, not the least of which was its conclusion that a “common language” already existed to a considerable extent in the use of an individual’s social security number as a “standard universal identifier.” Because almost every American had one, the number was much more reliable than, say, a person’s given name or driver’s license number. The HEW report opposed the establishment of any standard universal identifier in the United States “now or in the foreseeable future.”^{Footnote 23}

The report was an important marker in that it was among the first of a handful of government statements that looked to define informational privacy protections on a broad level and the principles it outlined clearly defined certain “non-trivial” privacy interests. This is significant when we consider that the United States, unlike most nations, never adopted an omnibus information privacy law that applied to the private sector. But as privacy experts point out, the report never questions the purposes for which data is collected nor does it provide an adequate explanation for when and how individuals might be able to simply “opt out” of data systems.

The timing of the HEW report was fortuitous in that it so shortly preceded the Watergate scandal. In many ways the Watergate scandal was about transparency, or a lack of privacy concerning institutions. But when Americans learned that the Nixon administration sought personal data collected by federal agencies to track and harass political enemies, many also came to believe that some regulation of that data was required to prevent future abuses of power.

Both houses of Congress set themselves to the task of addressing the growing public demand for tougher privacy legislation and eventually passed the Privacy Act of 1974. The bill drew heavily from the recommendations of the HEW report. In the hearings, many privacy advocates employed the familiar barrage of problems associated with information aggregation, insecurity, secondary use, and the growing reality of an inescapable past. Unlike the first round of hearings in the mid-1960s, they were able to draw from an array of horror stories thanks to newspaper stories and descriptive complaints submitted by constituents.

One anecdote from a Columbia Law Review article submitted to the Senate record illustrated perfectly the intersection between private and public data collection. An Army intelligence agent whose duties included security checks was ordered by a superior to conduct an investigation of the arrest, court, and bank records of a private citizen. He later learned that the investigation had been ordered by an intelligence officer as a personal favor to an executive of a private company. When the agent approached his superior, he was told the search was a matter of national security, but after reporting the incident to a pentagon official his suspicions were confirmed and disciplinary action was taken against the guilty officer. It was too late to help the subject of the search, though, who by then had already been fired.^{Footnote 24}

Proposals raised by privacy advocates in both houses met with considerable resistance from various interests who profited from information collection. Many argued that the dangers of data-related privacy invasions were too vaguely defined and that Congress was essentially gearing up to legislate on a problem it still did not understand. This line of attack was complimented with claim after claim that too much restrictive legislation would inhibit government efficiency and create an undue financial burden on private businesses. The arguments proved remarkably effective.

Interests like the American Life Insurance Association, representing 367 companies accounting for over 90 percent of the legal reserve life insurance in the United States, urged Congress to recognize a distinction between public and private information collection and stressed that any restrictions passed by congress should apply only to government agencies. Like most of their allies, they argued that “the proposed controls on private information systems” were entirely unrealistic—“so broad as to cover virtually every information practice and procedure that is a part of almost every business transaction,” and that their effects would be “extremely burdensome . . . not only unfair, but unworkable.” Two hundred sixty literary publishers argued that unless the direct-mailing companies and mail-order book clubs were explicitly exempted, “the impact on the publishing industry could be devastating.” The California Bankers Association noted a lack of proper scope, claiming the proposed bill would “apply to the ‘mom and pop’ corner grocery store as fully as it does IBM.”^{Footnote 25}

When the dust settled, both houses of Congress agreed on a compromise measure that became the 1974 Privacy Act. The law gave individuals rights of access and correction to computerized data files and placed restrictions on the collection, use, and disclosure of personal information by federal agencies. The act did not apply to records kept by state and local governments or those of private companies, and the Federal Privacy Board proposal never found its way into the final bill. President Ford deflected any private-sector provisions by establishing a Privacy Protection Study Commission to investigate the issue and recommend appropriate oversight in time.^{Footnote 26}

Contemporary legal scholars have been quick to point out the inadequacies of the Privacy Act, most notably its glaring lack of regulation over the private sector. Priscilla Regan believes it “encompassed the minimum protection that was advocated at the time” and “would not have been seriously debated and passed in 1974 without the Watergate revelations.” Its failure to create a privacy board kept the law’s implementation bogged down in questions of application. Despite a provision intended to regulate the use of Social Security numbers it did little to curtail their use as a common universal identifier. It also contained a rather large loophole with a “routine use” exemption, which allowed the disclosure of personal information by federal agencies if the purpose was compatible with the purpose for which it was collected.^{Footnote 27}

So why was the final bill’s reach so limited? The arguments of the data interests were certainly effective, but much of the blame falls on the privacy advocates themselves. By framing privacy again and again as an “individual right,” they ignored larger social considerations and left themselves exposed to charges of myopia.

Privacy advocates failed to position privacy as a larger societal right essential to the progress of a liberal civilization. Regan regards this limited conceptualization as a chief detriment to the formulation of effective privacy law, noting that “an individualistic conception of privacy does not provide a fruitful basis for the formulation of policy to protect privacy. . . . [We need] a logic for a new way of thinking about privacy—not solely as being important to an individual but also as having broader social importance.” Nissenbaum agrees, noting that “we need an account of privacy that justifies its place among shared collective values of a community so that privacy is accepted as a legitimate reason for accepting or rejecting a given socio-technical system, even by those members whose interests are not best served by the conclusion.”^{Footnote 28}

It is also important to note that in the short time between the data center debates and the 1974 Privacy Act, information-processing systems were becoming more democratized and information itself more mobile and profitable. When lobbying against the data center in the mid-1960s, privacy advocates were essentially squaring off against an archetypal force that was singular and identifiable. As such, they were operating within well-established parameters in the American social and cultural tradition. By 1974 privacy advocates were pushing against a much more varied array of interests comprised of multiple actors. The rhetoric employed by these interests, particularly their economic arguments, complicated the terms of the privacy debate.

The shortcomings of the 1974 Privacy Act are, in a way, quantifiable. A decade after its passage, the literal 1984, incidents of data sharing between private and public agencies reached into the billions—thanks largely to the advent of federally sanctioned data mining, computer matching, and the Reagan administration’s pursuit of “smaller government” under the banner of bureaucratic efficiency.

“Data mining” compares two or more sets of digital records (or “cross referencing”) to search for individuals who may be included in more than one file. A government agency might check federal employment records against, say, welfare records in the Chicago area like it did in 1976. An agent keys in the parameters of the search on a computer, presses enter, and in a few moments a mainframe spits out what those in the matching business call “raw hits.” The Chicago case produced 4,600 raw hits, or instances of government employees who were potentially receiving welfare benefits improperly. That number is then whittled down by operatives who check for errors and other mitigating factors that would exempt someone from the large list of suspects. After a round of vetting, those who remain on the list are dubbed “solid hits” and their contact information is passed on to the FBI. Hundreds of state and local prosecutions follow. Some are exonerated for reasons not found in the initial vetting. Some are caught swindling the system and convicted of fraud.^{Footnote 29}

Data matching gained a reputation as a money saver that enhanced bureaucratic efficiency, and throughout the 1980s states implemented their own matching programs. Over time, searches started bringing private institutions into the fold—particularly large employers and banks. Between 1980 and 1984, federal and state agencies conducted more than 533 computer matches that touched on more than 7 billion individual records. In that span, the frequency of searches nearly tripled.^{Footnote 30}

Computer matching meshed well with the policy rhetoric of the conservative ascendency of the 1970s and 80s, and proved useful weapons in that movement’s crusade to eliminate bureaucratic waste and the “cancer” of inefficient “big government.” Soon after moving into the Oval Office, President Reagan established the President’s Council on Integrity and Efficiency (PCIE) to enhance cooperation between federal agencies toward reducing fraud and waste. The council pushed hard for the establishment of a long-term computer-matching project and began developing federal inventory of computer-matching software packages.^{Footnote 31}

On the surface, it might appear that data matching lined up just as well with supporters of “big government” as it did with fiscal conservatives. It was, after all, making the welfare state more efficient and less prone to fraud. But it’s important to understand the social and political ideologies of those actors who were most responsible for implementing computer matching on a massive scale in this particular historical moment.

Data mining spawned largely from a political movement that made dismantling many of the social welfare aspects of “big government” a cornerstone of its legislative agenda. The New Right, as some call it, was hardly uniform. But in terms of socioeconomic policy, as numerous historians have noted, the Republican Congress of the 1980s campaigned on drastic cuts in public assistance, food stamps, environmental agencies, and tax cuts, while calling for a “New Federalism” aimed at reducing the size and scope of the federal government and at returning regulatory authority to the states. “Government,” President Reagan said again and again, “is not the solution, it’s the problem.”^{Footnote 32}

One of the larger computer-matching programs of the Reagan years was the ominously named Project SPECTRE,^{Footnote 33} which matched Social Security benefit payments against Medicare death records. The HEW reported that the search led to $7.5 million in recoveries to the Social Security trust funds and total of $25.3 million in overall savings. Of the original 8,300 raw hits, 5,263 persons were found to have been collecting benefits improperly assigned to dead people. Most payouts were simply terminated. Criminal indictments were brought against sixty-two individuals suspected of willfully deceiving the federal government, of which twenty-four received felony convictions.^{Footnote 34}

As the literal 1984 loomed, many press outlets throughout the country began running stories about computer matching and its privacy implications. Those with a penchant for irony were quick to note that a vast expansion of federal information practices was being conducted under the banner of “small government.”

A new series of congressional hearings to address the privacy concerns of computer matching convened in the early 1980s with numerous privacy advocates on hand to voice their objections. The OTA noted that there were “numerous procedural guidelines for computer matching, but little or no oversight, follow-up, or explicit consideration of privacy implications,” and that “Congressional actions seem to be contradictory” in “placing restrictions on agency disclosure of personal information” while subsequently endorsing computer matching. Most often, the subjects are federal employees and welfare recipients.^{Footnote 35}

The ACLU’s John Shattuck was particularly vocal in this period, arguing before the Senate that “once a computer match has taken place, any person whose name appears as a raw hit by virtue of turning up on two unrelated computer tapes is often presumed to be guilty of the wrongdoing that is under investigation.” Shattuck related a number of anecdotes to illustrate the potential dangers of information sharing to the individual citizen. One involved a middle-aged New Yorker who was denied a taxi license because a computerized report showed that when he was thirteen years old, in Massachusetts, he was temporarily placed in a mental institution. What the files did not show was that he was an orphan and the institution was the only home state authorities could find for him because of overcrowding.^{Footnote 36} “In the last 5 years,” argued Shattuck, “I think the [Privacy] Act has been so thoroughly circumvented by the executive branch that, frankly, it can no longer be seen as an effective safeguard against computer matching abuses.”^{Footnote 37}

Privacy advocates also attacked the efficiency argument by calling attention to the fact that many state and federal agencies used confusing methods to measure the savings yielded by computer matching—that the actual fiscal benefits were still unclear. A 1986 OTA report did indeed find there was “no firm evidence . . . to determine the costs and benefits of computer matching.” Agencies never operated under uniform standards of fiscal measurement and there were a number of dynamics at play, many of which are very convoluted.^{Footnote 38}

These arguments, similar to those made in Kenneth Laudon’s 1986 book, Dossier Society, did little to curb the expansion of federal computer matching between agencies and into the private sector. The efficiency line proved markedly effective, while privacy problems, however legitimate, were too vaguely formulated as a pressing societal concern.^{Footnote 39}

Advocates of computer matching were able to make the case for a larger societal good where privacy advocates had failed to do so. For every taxi driver denied a license, proponents of matching pointed to droves of welfare cheats—the very real people who undermined programs that were enacted in the spirit of the public good, like government employees who were delinquent on child-support payments and doctors who received Medicaid and Medicare payments after defaulting on their federal student loans. These cases became useful symbols for the Reagan administration’s political rhetoric of waste, the need for small government, and “welfare queens” in “gold Cadillacs.” In their own way, those legitimate offenders caught by computer matching contributed to weakening of American privacy.

Congress did pass the Computer Matching and Privacy Protection Act (CMPPA) in 1988, which required government agencies “to formulate procedural agreements before exchanging computerized record systems” for purposes of computer matching. But in retrospect, many privacy advocates viewed the law as sort of a Trojan horse. Computer matching was finally regulated, but the law mainly reflected issues of cost-benefit, due process, and administrative efficiency—not privacy. It also effectively legitimized computer matching as a desirable part of American life, institutionalized the routine sharing of information between agencies, and created the de facto national data center privacy advocates had opposed for decades.^{Footnote 40}

Because there is no tradition of studying information privacy within the discipline of United States History, any article-length glance at the larger narrative of pre-Internet data collection is inherently limited. Such a thing can only aim to flesh out a conversation. But if historians do choose to engage information privacy, one of the paramount issues of our age, they must do so with an understanding that data collection in the twentieth century responded to both liberal and conservative political currents; that it became increasingly democratized over time, which in turn made the terms of the surrounding debates increasingly complex and harder for privacy advocates to push against; and that while the nature of its regulation provided some protections, those regulations also served to legitimize the need for institutions to collect vast amounts of information about individuals while never challenging the basic assumptions behind these practices.

Historians would do well to familiarize themselves with the vast literature on privacy contributed by other disciplines in the social sciences, many of which offer a number of useful frameworks for historical engagement. The goal, as with other important social questions, is a symbiotic relationship where historical studies of privacy would both draw from and contribute to other perspectives on this issue.^{Footnote 41}

One place historians might lend perspective is an examination of how the profit motive, something history shows us has traditionally been at odds with the aims of privacy advocates, is today being utilized as an incentive to protect individual privacy. To be sure, private companies still make billions from data collection, but in the wake of the Edward Snowden revelations many in the business community have begun catering to consumers who want their personal data protected. An increasing number of search engines, content-management services, mobile phone providers (like Apple), and social-networking companies are beginning to aggressively market themselves as entities that will not share their customers’ data with outsiders. Privacy, in this respect, is fast becoming a commodity, and the profit motive is no longer exclusive to only one side of the debate. This seems to be a crucial historical shift.^{Footnote 42}

As the discourse over privacy continues, the United States will likely do what it usually does: reexamine its past and drag center stage historical developments that have been previously glossed over. Each side in the coming debates will likely have its champions, well studied in sophistry and playing down details unsupportive of their agendas. It is the duty of scholars to provide perspective here by stressing relevant complexities and dishing out a larger, less rhetorical view.