The Disparate Impact of Surveillance

The Disparate Impact of Surveillance

Mass surveillance society subjects us all to its gaze, but not equally so. Its power touches everyone, but its hand is heaviest in communities already disadvantaged by their poverty, race, religion, ethnicity, and immigration status. Technology and stealth allow government watchers to remain unobtrusive when they wish to be so, but their blunter tools—stop-and-frisk, suspicionless search, recruitment of snitches, compulsory questioning on intimate subjects—are conspicuous in the lives of those least empowered to object.

We are not exactly the first to notice these disparities. There is a rich and diverse literature of dissent, hard experience, and scholarship about the disproportionate intrusions of government into poor and brown communities, much of it produced by people with roots in those precincts themselves. We are indebted to their work. We will make no pretense here of explaining the unjust burdens of surveillance to people who have carried and protested them for decades. We direct our report, instead, to people who could have listened but did not: lawmakers, law enforcement authorities, mayors, governors, and other people with power who failed to look beyond their own experience and point of view. Most of all, we aim this report at our fellow thinkers about surveillance on a national and global scale.

For decades, and especially in the past four years, civil libertarians (including the authors of this report) have tended to frame this debate in the discourse of universal rights. We assert a place for privacy among the core liberties that restrain state power in a self-governing democracy. Through that lens, government surveillance is seen as inflicting its harms on everyone. Even when ostensibly targeted, its methods tend to encroach on the public at large. All of us, consciously or not, have something to hide—something that would harm us or people we know if revealed in the wrong circumstance. We stand by these positions. We have also come to see them as profoundly incomplete.

Universalist arguments obscure the topography of power. Surveillance is not at all the same thing at higher and lower elevations on the contour map of privilege. Privacy scholars speak of philosophical rights and hypothetical risks; privacy-minded middle class Americans fear allowing the government too much access to their electronic trails. But there is nothing abstract about the physical, often menacing, intrusions into less fortunate neighborhoods, where mere presence in a “high-crime” area is grounds for detention, search, and questioning by police. At age sixty-five, tens of millions of Americans claim their Medicare benefits with nothing more eventful than completion of some forms. (Medicare.gov even promises to “protect your privacy by getting rid of the information you give us when you close the browser.”) An impoverished single mother on Medicaid faces mortifying questions, face-to-face with benefit managers, about her lovers, hygiene, parental shortcomings, and personal habits.

The abusive application of broad legal powers should color our choices about the powers themselves. We do not mean “abuse” in the narrow sense of a rogue employee who deliberately breaks the rules. Much more common and more serious problems arise from officially sanctioned practices—ostensibly within the rules—that authorities inflict unequally on poorer, browner, and blacker communities. State and local overreach of this kind should be no less central to the national privacy debate than intelligence operations on a global scale.

Privacy is not only a luxury that many residents cannot afford. In surveillance-heavy precincts, for practical purposes, privacy cannot be bought at any price.

This report gives primary attention to the disparate impact of surveillance made manifest through urban policing and social services for the poor. Not everyone is accustomed to viewing these systems through the lens of surveillance, but a substantial body of research has done so for years. In the lived experience of minority neighborhoods, authorities at ground level wield enormous power to peer into their lives. Privacy is not only a luxury that many residents cannot afford. In surveillance-heavy precincts, for practical purposes, privacy cannot be bought at any price. Unspoken policy, well intended or not, disqualifies those people from basic protections that most Americans can take for granted.

Privacy advocates have sometimes struggled to demonstrate the harms of government surveillance to the general public. Part of the challenge is empirical. Federal, state, and local governments shield their high-technology operations with stealth, obfuscation, and sometimes outright lies when obliged to answer questions. In many cases, perhaps most, these defenses defeat attempts to extract a full, concrete accounting of what the government knows about us, and how it puts that information to use. There is a lot less mystery for the poor and disfavored, for whom surveillance takes palpable, often frightening forms.

Universalist arguments apply across lines of social and economic class, but their generality is not well suited to combating unequal and particular harms. Members of our most disadvantaged communities have long asserted that they fall under heavier and more hostile scrutiny, and empirical research supports them. Meaningful oversight of state surveillance cannot be designed with only common conditions in mind.

There is no sense in having a conversation about surveillance in America without taking account of its profoundly unequal use. We cannot divorce privacy research or policy from considerations of class, race, gender, and other social hierarchies. When we do that, as we do often, we fail. The liberal conception of privacy has no claim to universal reach if it fails to adapt to unequal circumstance.

The Birth of “Privacy”

The concept of privacy was conceived in counterpoint to the government’s growing ambition to peer inside long-closed compartments of our personal lives. The struggle took place most often in court. The leading cases mark milestones in that history, and show the disproportionate place of minority surveillance in the evolution of law.

The provenance of privacy as a legal right is generally traced to Louis D. Brandeis. In “The Right to Privacy,” published in the Harvard Law Review in 1890, Brandeis and Samuel D. Warren, his law partner and former classmate, made the case for a “right to be let alone” embedded in English Common Law. Seventy-five years passed before the Supreme Court gave its blessing to a version of that claim.

In 1928, by then an associate justice of the U.S. Supreme Court, Brandeis wrote an influential dissent in Olmstead v. United States that reframed the right to privacy in constitutional, not Common Law, terms. The majority in Olmstead upheld a warrantless wiretap because police had not trespassed on the defendant’s property or searched his “tangible material effects.” In his dissent, Brandeis wrote that “material things” and places were not the main concern of the Founders. “They sought to protect Americans in their beliefs, their thoughts, their emotions and their sensations,” he wrote. “They conferred, as against the government, the right to be let alone—the most comprehensive of rights and the right most valued by civilized men.” An “unjustifiable intrusion,” whether or not by physical trespass, breached the Fourth Amendment. The use of evidence so obtained in a criminal case, he wrote, “must be deemed a violation of the Fifth.”

Nearly four decades later, in 1965, the Court cited the Olmstead dissent when it struck down a state ban on contraception for married couples. Griswold v. Connecticut recognized a Constitutional right to marital privacy in the “penumbras” of the Bill of Rights, glimpsed only in the shadows of the text itself. Griswold, in turn, gave rise to a line of cases that rested the right to privacy in the liberty interest protected by the Due Process Clause of the Fourteenth Amendment, which enabled the court to expand its authority over state law. Eisenstadt v. Baird extended the holding in Griswold to unmarried straight couples in 1972. Roe v. Wade, the following year, used the same reasoning to protect a woman’s right to choose an abortion. It took thirty more years for the court to recognize the same basic freedom in the private lives of lesbian, gay, bisexual, and transgender people. Not until 2003 did Lawrence v. Texas extend Griswold far enough to strike down a Texas anti-sodomy law. The Court found “no legitimate state interest which can justify an intrusion into the personal and private life of the individual.”

Taken together, the Court’s rulings established a constitutional right to privacy in family, marriage, parenthood, procreation, and sex. This right—sometimes termed “decisional privacy” or “personal autonomy”—was subject to balancing against other “compelling interests” of the state.

The Court found another source of privacy rights in the First Amendment. Stanley v. Georgia, in 1969, struck down a Georgia law forbidding the private possession of obscene materials. “If the First Amendment means anything,” wrote Justice Thurgood Marshall for the majority, “it means that a State has no business telling a man, sitting alone in his own house, what books he may read or what films he may watch. Our whole constitutional heritage rebels at the thought of giving government the power to control men’s minds.”

The U.S. Supreme Court has declined to recognize “informational privacy,” or control of information about oneself. Legislators have done so, selectively, by statute. Between 1970 and 1999, Congress passed a series of laws restricting the collection, retention, sharing, and use of certain categories of non-public information. The laws governed credit records, “personally identifiable information” held by government agencies, financial information in the hands of commercial or investment banks, and non-anonymous information about children under age thirteen.

Over half a century, collectively, Congress and courts granted limited protections against unwanted observation, unwanted search, and unwanted control of private behavior. In each of those spheres, we will argue, the growth of state power has now overwhelmed the defenses—most starkly, and most unjustly, in marginalized communities.

The Color of Surveillance

Privacy as a matter of law progressed along multiple paths, some poorly marked, with no obvious destination in common. Its meanings are diverse. We have not seen a persuasive attempt to write a single definition of “privacy” that covers physical search and seizure, confidential communications, and the “freedom to be let alone” in personal behavior that Brandeis described as privacy’s core.

We do not need a unified theory of privacy to show that, in each of its meanings, marginal communities enjoy far less of it in practice.

We do not need a unified theory of privacy to show that, in each of its meanings, marginal communities enjoy far less of it in practice. In some contexts, poor people and people of color have legal rights to privacy, but no means to exercise them; “paper rights,” as Karl Llewellyn called them. In other contexts, the government justifies extraordinary surveillance in superficially general language that applies exclusively, or close to exclusively, in minority neighborhoods. In still others, the government denies a disfavored class a privacy right, even in principle, that other Americans freely enjoy.

We focus most concretely in this report on excesses of surveillance in policing and in the interactions of the poor with the welfare state.

Policing

Surveillance in America owes its origins, in part, to the slave economy. Poet and media activist Malkia Cyril puts the point provocatively: “From colonial times to now, surveillance technologies have been used to separate the citizen from the slave, to protect the citizen from the slave.” Plantation ledger books served as proto-biometric databases, recording the slaves as physical specimens in fine detail. The slave pass, the slave patrol, and the fugitive slave poster—three pillars of information technology in their day—prefigured modern policing, tracking, and photo ID. Plantation space, by one historian’s account, was organized to enable planters and overseers to “exercise surveillance and reinforce the subordinate status of enslaved people.”

In response to a slave revolt in 1712, New York City passed a law requiring slaves to carry lit lanterns if they were out after sunset without the company of a white person. The intent was to ensure that the enslaved population could be seen, located, and controlled at all times. As UT–Austin sociology professor Simone Browne has observed, light and sight remain key “disciplinary tactics” in the state’s arsenal.

In 2014, the New York Police Department, backed by a liberal Democratic mayor, deployed massive, blinding floodlights in minority neighborhoods and pointed them at public housing projects in the name of deterring crime. Heavy police reinforcements arrived, blue lights flashing all night, to patrol residential towers. Mayor Bill de Blasio said the goal was to convey “omnipresence”—a visceral message that police were everywhere. In a population that had lived through years of the previous police campaign of stop-and-frisk, the new one felt like an occupation. For poor communities of color, omnipresence was old news.

Money, Class, and the Distribution of Privacy

The Fourth Amendment, among our strongest safeguards against police overreach, arose in part as an expression of property rights. One guiding principle in English Common Law was that “home is man’s castle.” Something resembling sovereignty prevailed within its walls. Before the state could breach the castle, the Founders chose to require a warrant based upon “probable cause, supported by Oath or affirmation, and particularly describing the place to be searched, and the persons or things to be seized.” Their focus on “houses, papers, and effects” linked privacy and private ownership from the start. In doctrine and practice, that connection still shapes the boundaries of the Fourth Amendment’s legal shield against unreasonable search and seizure.

In obvious and less obvious ways, wealth plays a role in defining who and what the Fourth Amendment protects. Apart from the Brandeis dissent noted earlier, the Olmstead case was an unexceptional affirmation of precedent. Chief Justice William H. Taft wrote for the majority that the Fourth Amendment had no bearing on searches of places and things outside a defendant’s home or property. The government required no warrant to eavesdrop on calls made by Roy Olmstead, an ex-cop turned bootlegger, because the telephone wires “are not part of his house or office, any more than are the highways along which they are stretched.”

Decades later, in another wiretap case, the U.S. Supreme Court shifted course. In Katz v. United States, decided in 1967, the Court embraced Brandeis’s argument that a search need not involve physical trespass to cross a constitutional line. Authorities had eavesdropped on the defendant at a telephone booth, not on his property, but the court ruled he had a “reasonable expectation of privacy” nonetheless. Justice John Marshall Harlan addressed the obvious ambiguity of that phrase in a concurring opinion. Harlan proposed two tests: “[F]irst that a person have exhibited an actual (subjective) expectation of privacy and, second, that the expectation be one that society is prepared to recognize as ‘reasonable.’”

Katz did not stray as far from Olmstead as first appeared. Privacy and property were still entwined, if less directly, by the nature of the factual question that Harlan asked. Could a reasonable person in the defendant’s shoes expect to stay out of earshot at the time and place of the wiretap? If so, surveillance without a warrant was analogous to peering inside a private home. Katz made his case successfully because he placed his illegal gambling calls in a telephone booth with a door that closed and glass walls that allowed him to keep an eye on surroundings. If a nearby stranger could have listened in, on the other hand, police could do the same on their own authority. The Katz doctrine in practice came to mean that “privacy follows space,” not citizens, as late legal scholar William Stuntz observed.

Unobserved space costs money. A class-statified spectrum of legal privilege took shape in the cases that followed Katz. Trial and appellate courts found, not unreasonably, that citizens had the strongest expectations of privacy in their homes, somewhat less in their cars, and not much at all on public streets or public transport, however intimate the words they exchanged with a seatmate on the bus. A nosy passenger behind them might overhear, and government authorities therefore felt free, without contradiction yet from the courts, to place hidden microphones under the seats.

Class-inflected hierarchies did not stop there. White-collar workers with office doors had legal rights against warrantless surveillance that did not apply on the shop floor or a construction site. Property ownership did not formally define a person’s Fourth Amendment rights, but a fenced-in yard, a house set back from the street, a bit of “curtilage” at the property line, or even curtained windows might decide whether Fourth Amendment protections applied. Thin-walled apartments in densely packed towers made for porous shields at best. The poor have little choice but to live out their private lives in public spaces: parks and courtyards, empty lots, sidewalk stoops. The homeless, with still fewer sanctuaries, have correspondingly even weaker legal claims to a “reasonable expectation of privacy.”

Together, these precedents have created an unequal distribution of privacy rights. The richer you are, the more privacy you enjoy; the poorer, the less.

In 2012, the Court adjusted course again. By unanimous vote it reversed a verdict against an alleged drug dealer convicted with tracking data from a GPS device that police placed surreptitiously on his car. The Justice Department, following precedent, maintained that anyone might have seen the defendant as he drove public streets and thus the FBI was free to track him too. In United States v. Jones, the Supreme Court disagreed, ruling that electronic surveillance of this sort counted as a “search” under the Fourth Amendment and therefore required a warrant based upon probable cause. The Justices disagreed, however, on why. The only argument that carried a majority was that, as Justice Antonin Scalia put it, police had failed “the common law trespassory test” by sneaking onto the defendant’s driveway for access to his car. In that respect, despite hints of a broader doctrine to come, Jones reinforced the old property-bound conception of privacy. Carpenter v. United States, which went to oral argument on November 29 of this year, presents a similar set of facts without a trespass on private property. (Police tracked the defendant by his mobile phone’s contact with cell towers.) The court’s decision, now pending, has the potential to set a new doctrine for the Fourth Amendment in the electronic age.

The Poverty Exception

Policy has to comply with law, but it does much more than that. Policy makers are free to consider facts and values that judges do not. They may use policy to correct what Vanderbilt University law professor Christopher Slobogin calls the “poverty exception” that courts have carved out of the Fourth Amendment. Something is very wrong when protection against “unreasonable search and seizure” depends on conditions that money can buy and the poor cannot hope to afford.

Stuntz hypothesized that poor neighborhoods are over-policed because the police face fewer constitutional limits there. Search and seizure and, therefore, arrest, are simply cheaper in budget and personnel. Considering the example of drug markets, Stuntz writes, “In well-off neighborhoods, transactions are likely to take place in private dwellings through arranged meetings; in poorer neighborhoods, transactions take place on the street. Fourth Amendment law makes it much harder to police the former, and thereby pushes police to focus ever more on the latter.”

That explanation fits bureaucratic behavior in other contexts. But it does not seem likely to cover all, or even many, of the motives for law enforcement scrutiny in disfavored minorities. The causal relationship, in many respects, could as easily be reversed. The law found more leeway to invade the Fourth Amendment rights in poor communities because the poor—especially the non-white poor—are perceived as uniquely menacing. “The police are not picking on the urban poor because the rest of society is too hard to search,” Slobogin writes. “[T]hey are simply going after the group they think, rightly or wrongly, is most crimogenic.” Regardless of motive, poor communities of color are policed not only more frequently but with a far heavier hand than whiter and wealthier ones.

The New York Police Department offered a powerful demonstration of that dictum in its eight-year program of stop-and-frisk in minority neighborhoods. From 2004 to 2012, the department subjected 4.4 million New Yorkers to brief detention, questioning, and (more often than not) a compulsory hands-on search. Along the way to finding the program unlawful, a federal judge forced the city to disclose that over 80 percent of the people stopped and frisked were black or Latino. Police found no weapon in 98.5 percent of the searches, conclusive proof in itself that the standard employed fell far short of “reasonable” and farther still of “probable cause.”

This was a critical moment of policy choice, not only a question of law. Even if courts had allowed it, the mayor and City Council had final say over the deployment and use of police authority on this scale. The Supreme Court had made exactly that point in Terry v. Ohio, the 1968 case that prefigured the New York program. The court ruled that year for the first time that brief detention and search on a public street did not fall within the Fourth Amendment’s scope. Police needed only “reasonable suspicion,” not probable cause to believe, that their subject was involved in a crime.

The Terry Court added a remarkable warning about the racial dimensions of mass detention and search. “In many communities, field interrogations are a major source of friction between the police and minority groups,” the court wrote, citing findings from a presidential commission. A frisk was no “petty indignity,” but a “severe, though brief, intrusion upon cherished personal security, and it must surely be an annoying, frightening, and perhaps humiliating experience.” The humiliation might even be deliberate, directed at “anyone who attempts to undermine police control of the streets.”

The justices, writing unanimously, were openly inviting political authorities to consider motivations and consequences that judges could not properly take into account. If the court intended those words as an implicit call for restraint, the explicit terms of its ruling had far more impact. Terry and its progeny empowered police to build an arsenal of intrusive techniques, with little risk of adverse rulings in court. In 1996, the Court extended Terry’s logic from pedestrians to motorists. Police could stop and search a driver for the smallest traffic violation—even if a reasonable officer would not do so. The latter point came close to acceptance of pretext, so long as any infraction could plausibly be claimed. The statistical evidence of racial profiling led critics to describe the true offense in these cases as “driving while black.” Four years later, the court held that a person’s mere presence in a “high crime area,” combined with “nervous, evasive behavior,” sufficed as grounds for reasonable suspicion.

Both factors supplied easy templates to officers on the front lines. NYPD officers cited them reflexively. On the UF-250 forms they had to fill out after every stop-and-frisk, the two grounds for suspicion they cited most often were “High Crime Area” and “Furtive Movements.” The first justification was based upon a largely subjective label that counted loitering and disorderly conduct but not, for example, fraud and other white collar crimes. It turned out to be hard to distinguish, statistically, from “high-black areas.” Fear and retreat—the second justification—described the well-grounded preference of minorities, for obvious historical reasons, to avoid contact with police.

The personal experience of stop-and-frisk, as the Terry court seemed to anticipate, could be brutal. A 2012 study, based on voluminous interviews, recounted episodes of “being forcibly stripped to their underclothes in public, inappropriate touching, physical violence and threats, extortion of sex, sexual harassment and other humiliating and degrading treatment.” One interviewee said, “It made me feel violated, humiliated, harassed, shameful, and of course very scared.” Because the stops took place in public and “everybody’s looking,” another subject said, “people get the wrong perception of you.” New Yorkers who lived in target neighborhoods often reported that they withdrew from social interactions, afraid to sit with friends on an outdoor porch or walk to a nearby store.

Many said they felt obliged to carry postal mail to prove they lived nearby and pay stubs to demonstrate a legitimate source of money in their pockets. Manny, a twenty-three-year-old Filipino- and African-American man in the Bedford–Stuyvesant neighborhood, said, “In my community, and a lot of communities in Brooklyn, it’s kind of like an occupation.”

As we debate surveillance in other contexts, thoughtful policy advocates should look to this reality for lessons on the risk of granting authorities too much discretion to intrude into private lives. Extraordinary powers meant for limited use become commonplace. Worthy goals bring unintended harm, and unworthy intent goes unseen. Temporary policy grows permanent by degrees. For all kinds of reasons, the minimum constraints of law are a grossly insufficient guide to boundary-setting by policy makers, political leaders, and citizens.

For all kinds of reasons, the minimum constraints of law are a grossly insufficient guide to boundary-setting by policy makers, political leaders, and citizens.

The minimum constraints may not even do much constraining. Despite Judge Shira Scheindlin’s ruling in Floyd v. City of New York, the NYPD maintains to this day that the stop-and-frisk program is a lawful and effective tool—despite ample evidence that it is not. The department has reduced but not halted the practice. In one variation or another, stop-and-frisk remains a standard tactic of big-city policing across America.

Privacy and Public Benefits

In June 2000, Rocio Sanchez separated from her husband and applied for Temporary Assistance for Needy Families, better known as “welfare,” to support herself and her infant daughter. One month later, an investigator from the Public Assistance Fraud Division of the San Diego District Attorney’s Office knocked on her door, unannounced. Sanchez was not personally suspected of fraud or ineligibility—she had yet to receive a check. But San Diego County had instituted a program called “Project 100%,” which required home inspections for every welfare applicant. The investigator asked Sanchez in intimate detail about her husband, their relationship, and the reasons for their split. He proceeded to search her home, interrogating her about a man’s shirt and shoes found in the closet, which belonged to Sanchez’s brother. The investigator went on to question her neighbors. When he encountered her again at her old apartment, where her husband had been living, he grew hostile and suspicious. She was cleaning up the apartment in order to reclaim her security deposit, but the inspector conducted another meticulous search and demanded that Sanchez pull papers from the trash and translate them. Submission to this stranger, who held her fate in his hands, was the price of approval for welfare benefits two months later.

If this assault on personal boundaries was not “surveillance,” we are defining it wrong. For many poor Americans, the cadres of the welfare state are the primary face of government. From the moment they apply for food stamps, Medicaid, public housing, or welfare, the poor fall into the domain of super-empowered snoops who cannot be turned away. Welfare applicants are subject to physical search, unbounded questioning, a presumption of cheating, and coercive rules that restrict their choices on such things as whether and when to bear children, and how many. Courts have justified these intrusions in the name of child protection, fraud detection, taxpayer savings, prevention of drug abuse, and promotion of whatever policy governs family planning at the time.

In an honest policy debate, beyond the narrow confines of a judge’s review, these claims would fail on their faces. Any government program with comparable benefits—Medicare payments, adoption assistance, dependent care savings accounts, mortgage interest deductions, or education loans—could claim a high-minded purpose for intrusive questioning. None inflicts routine abuses of this kind on citizens in the financial and racial mainstream. Boston University legal scholar Khiara Bridges attributes the differences to “the moral construction of poverty”: the idea, consciously or unconsciously held, that poor people are poor because of character failings that pose a threat to themselves, their children, and society.

The policing and welfare functions of the state are closely entwined. Both belong to a broader project of discipline and control. “Welfare reform” in the 1990s brought a bipartisan shift toward punitive approaches to public assistance. Policy vernacular at the time called for “less carrot, more stick.” Block grants to states funded “temporary assistance” (the kind Sanchez asked for), abandoning the more generous terms of Aid to Families with Dependent Children. Welfare bureaucracies built an elaborate system of sanctions for adults who failed to make the transition from welfare to work, with more onerous behavioral checks and a more aggressive approach to fraud detection. As a result, writes UC-Irvine professor of law Kaaryn Gustafson, “The welfare system and the criminal justice system in the United States are becoming ever more tightly interwoven.”

“Deepest Legal Intrusion”

The Rocio Sanchez story became a leading case in welfare surveillance law. The result showed once again that public policy on search and seizure should not accept the most aggressive surveillance, on the least plausible grounds, that a court will tolerate.

In 2006, the Ninth U.S. Circuit Court of Appeals found no violation of Sanchez’s Fourth Amendment rights. Home visits by welfare agencies, by previous U.S. Supreme Court precedent, did not count as warrantless searches. In the first place, Sanchez was not formally “forced or compelled” to submit. She was free, the state reasoned, to refuse entry to her home and find some other way to feed her daughter. In the second place, the state declared a “rehabilitative” purpose. The Sanchez case, however, was not even superficially about rehabilitation. A fraud investigator sent by the San Diego County District Attorney’s Office, not a welfare case worker, searched her home.

In its most revealing passage, the Ninth Circuit panel’s opinion proposed an analogy. A criminal on probation, the court noted, has a “relationship with the state” that reduces his expectation of privacy from a probation officer in his home. Sanchez, by applying for benefits, had placed herself in a comparable relationship with welfare investigators.

Even without that revealing conflation of welfare and crime, the court’s depiction of facts was entirely disconnected from reality on the ground. Sanchez had given “uncoerced” permission to search her home, the court reasoned, because denial of her legal right to benefits did not count as a direct penalty for refusal. Regardless of its formal merits, this argument showed remarkable indifference to the power relationships and the stakes for Sanchez. No young mother would pause for a moment’s thought if asked to decide between her Fourth Amendment rights and subsistence for her child. Judge Harry Pregerson, speaking for eight dissenters when the Ninth Circuit declined to rehear the case en banc, observed acidly that other Americans, outside the welfare system, are not typically asked to submit to “humiliation and further assaults on their dignity” when they apply for entitlements:

The government does not search through the closets and medicine cabinets of farmers receiving subsidies. They do not dig through the laundry baskets and garbage pails of real estate developers or radio broadcasters. The overwhelming majority of recipients of government benefits are not the poor, and yet this is the group we require to sacrifice their dignity and their right to privacy.

Too much blame on the courts is misdirected. It was the San Diego County District Attorney’s Office, not the appellate judges, who devised their cavalier program of hazing for welfare applicants. Perhaps the district attorney saw political advantage there. Furthermore, the Ninth Circuit ruling concerned a narrow question of law; it added no substance or moral force to San Diego’s mistreatment of the poor. When the public debate resumes—in more favorable circumstances, one hopes—advocates of reform owe no deference to the status quo because it passed legal muster with the court. There is room for a lot of bad policy under law. In the end, it is the job of political leaders and policy advocates to propose a more just and effective system of oversight.

There is room for a lot of bad policy under law. In the end, it is the job of political leaders and policy advocates to propose a more just and effective system of oversight.

Castles, Public and Private

Public and subsidized housing, no less than welfare, have historically been Fourth Amendment exclusion zones. In the early 1990s, as part of a self-described campaign against violent drug crime, police in Chicago, Baltimore, Philadelphia, and elsewhere commenced impromptu, warrantless raids of public housing complexes. Chicago’s “Operation Clean Sweep” involved twelve steps, including “securing perimeter of building by placing police at all entrances and exits,” “notifying the press that a sweep is under way,” “inspecting each unit,” and “enclosing the lobby to control access.” Many Chicago housing residents had grown weary of gang violence; they welcomed the raids. Others felt their rights were under assault. In 1994, a federal court deemed “Clean Sweep” unconstitutional on Fourth Amendment grounds. Judge Wayne Anderson acknowledged in his ruling, “Many tenants within CHA housing, apparently convinced by sad experience that the larger community will not provide normal law enforcement services to them, are prepared to forgo their own constitutional rights.” He was unwilling to permit the government to “suspend their neighbor’s rights as well.”

In response, Attorney General Janet Reno and Housing Secretary Henry G. Cisneros devised what President Clinton called a “constitutionally effective way” of conducting sweeps. “Every law-abiding American, rich or poor,” Clinton said, “has the right to raise children without the fear of criminals terrorizing where they live.” The president called for more patrols of common areas and vacant apartments, and for warrantless searches in “emergency” circumstances. Most notably, he called on city officials to add “consent clauses” to public housing leases, obliging tenants to grant blanket permission for police to enter their apartments without probable cause. Instead of imputing implied consent, as the Sanchez case allowed San Diego to do for welfare applicants, Clinton proposed to compel tenants sign away their Fourth Amendment protections on pain of eviction.

Few housing authorities adopted Clinton’s proposal, but the zone of privacy for tenants continued to shrink. Between 1991 and 2014, as part of the “clean halls” program, the NYPD routinely patrolled the public spaces and stairwells of housing projects to find and remove nonresidents, who were automatically declared to be “loitering,” once again an unthinkable policy in other precincts. Police were obliged to shelve “clean halls” in their 2014 settlement of a civil suit on stop-and-frisk. Now officers need “reasonable suspicion” of illegal presence—the Terry stop standard—to expel someone. Housing officials in Longmont, Colorado, came up with another dodge. In June 2017, they invited police to bring K9 drug teams for walk-throughs of low-income apartment complexes as “an opportunity for the dogs to train.

Recently, police in New York have used racketeering charges to justify massive, military-style raids at housing projects. Under the Racketeer Influenced and Corrupt Organizations (RICO) Act, police can conduct searches and arrests on the basis of patterns of activity within a criminal enterprise and allegations of criminal offenses for which police have no individually specified probable cause for arrest. The rules, designed for flexibility against the mafia, allow police to conduct searches and arrests in large numbers, deferring requirements for individual evidence of guilt, within the confines of a broadly described criminal enterprise.

Local cop-watchers have documented at least twenty such raids since 2014, many of which have terrorized residents and swept up hundreds of young men of color on suspicion of gang activity. “Taking people in the middle of the night like that, that doesn’t make us feel any safer,” said one twenty-six-year-old woman, whose Bronx housing complex, Eastchester Gardens, was the target of the city’s largest-ever gang raid last year. “They hit people with conspiracy, why?” says Marcus Wray, twenty-eight, a longtime resident of Eastchester Gardens. “Because when you hit someone with conspiracy you can get every person around. The person sitting on the bench. Because I know that person, I might get indicted with him.”

According to a University of Cambridge study of publically available RICO gang indictments, 86 percent involve gangs primarily composed of blacks, Latinos, and Asian Americans. By contrast, observes New York City journalist Simon Davis-Cohen, “The KKK is not treated like a gang or indicted on federal conspiracy charges.

Indeed, there is no neutral definition of a “gang” and the term has no special meaning as a matter of law. There are, of course, genuine criminal enterprises that could reasonably be described as gangs, from the Hells Angels to MS-13. Among the targets of special police scrutiny in minority urban neighborhoods, “gang” is more nearly a label than a documented description of fact. In some cases, it describes no more than a nameless group of young brown or black men who are suspected, collectively, of criminal acts or intent. (If there were probable cause to charge anyone in particular, he would be in custody.) The “gang” designation justifies the use of more aggressive surveillance and other police tools, with less attention to the niceties of constitutional restraint. Police departments catalog alleged gang members in massive databases without a semblance of due process. Tattoos of a certain kind, mere presence in a “known gang location,” or police observations that suggest friendship with another suspected gang member are enough to qualify.

Inclusion in that database can have dire consequences. Many states have enhanced sentencing for gang members. In a survey of sixty-four defense attorneys conducted by K. Babe Howell, a professor of criminal law at the City University of New York, 90 percent said gang affiliation had come up at bail hearings; two-thirds said their clients had received substantially higher bails for that reason; and 60 percent said they had clients who were denied bail altogether on the basis of inclusion in a database that authorities were not obliged to justify. According to Howell’s data, 90 percent of those added to the NYPD’s gang database between 2001 and 2013 were black or Latino. Just 1 percent were white. In most states and cities using gang databases, there is no process either for a resident to find out if he or she is on one of these lists or to offer evidence that police are mistaken. The link to racially tainted mass incarceration is quite direct. Every criminal lawyer knows that a defendant in lockup is overwhelmingly more likely to accept a guilty plea, rather than wait as much as a year behind bars for a trial.

Welfare Surveillance

Far beyond the San Diego example, suspicion and pervasive surveillance accompany state cash benefits for the needy. “Throughout the country,” writes University of Baltimore professor of law Michele Gilman, poor mothers are forced to “comply with extreme verification requirements to establish eligibility for welfare benefits, travel to scattered offices to procure needed approvals, reappear in person at welfare offices at regular intervals” and “answer intrusive questions about their child rearing and intimate relationships.” Comprehensive personal and financial data is collected, shared, and cross-checked against federal and state databases—all of which become accessible to police. An elaborate system of rules constrains the behavior of welfare recipients, and they must report minute changes of circumstance to welfare authorities. As a result, Ohio University professor of political science John Gilliom observes, many welfare recipients “live with surveillance as a totalizing and encompassing force.”

In his book Overseers of the Poor, Gilliom conducted an ethnography of fifty welfare-receiving mothers in rural Appalachia. Their words underscore the effect of the state’s constant scrutiny and suspicion on their well-being and sense of autonomy. Mary, a forty-year-old mother of three in Ohio, said that “You have to watch every step like you are in prison. All the time you are on welfare, yeah, you are in prison. Someone is watching like a guard. Someone is watching over you and you are hoping every day that you won’t go up the creek.” Jesse, a twenty-one-year-old mother of one, put it succinctly, “They act like they own us.” In Austin Sarat’s study of the “legal consciousness” of the welfare poor, many respondents describe a feeling of enclosure. “For me the law is all over,” said Spencer, a thirty-five-year-old beneficiary, “I am caught, you know. There is always some rule that I’m supposed to follow, some rule I don’t even know about.”

Gilliom maintains that these surveillance systems are not primarily a defense against fraud. They are meant to “control human behavior” and “enforc[e] rules and norms by observing acts and recording acts of compliance and deviance.”

These are, again, fundamentally political choices of policy—vouchsafed, perhaps, by federal courts, but ultimately susceptible to change. The discriminatory and dignity-denying scrutiny inflicted on the poor should be a primary concern for any self-described privacy advocate.

Your Psychosocial Examiner Does Not Work for You

Mothers and pregnant women on Medicaid, the low-income health care program, are equally held captive by an overwhelming system of scrutiny and control. Here again, this special burden has no analog in Medicare, the much larger health care entitlement for elderly and disabled Americans. (Medicare is an immense bureaucracy, with all the frustrations that entails, but it does not inflict routine abuses of the kind we discuss here.)

Boston University professor of anthropology Khiara Bridges has documented in infuriating detail the invasive, exhausting, and sometimes traumatizing prerequisites of an application. Bridges made a similar evaluation of New York State’s Medicaid-funded Prenatal Care Assistance Program. In order to enroll, before even seeing a doctor, midwife, or nurse, a pregnant woman must answer scores of questions that may not only embarrass her, but also expose her to risks she cannot control: questions about immigration status, illegal drug use, and domestic violence by a present or former partner (who could harm her again if he learns she has broken her silence). She has no practical way to discover, under pressure of time and medical need, how far and wide her answers will be shared.

During this period, the expecting mother undergoes a “psychosocial assessment.” Despite any appearance of assisting, the psychographic examiner is not employed on her behalf, and offers no assurance of confidentiality. Rather, the whole purpose of this review is to gather and record a catalog of risk factors, including:

the unplanned-ness and/or unwanted-ness of the current pregnancy; the woman’s intention to give up the infant for adoption or to surrender the infant to foster care; an HIV-positive status; a history of substance abuse; a lack of familial or environmental support; marital or family problems; a history of domestic violence, sexual abuse, or depression; mental disability; a lack of social welfare benefits; a history of contact with the Administration for Children’s Services [child protective services]; a history of psychiatric treatment or emotional disturbance; and a history of homelessness.

Bridges adds that all this information is “made into objects of knowledge for citywide, statewide, and nationwide statistics.” It is very, very hard to conceive, regardless of the professed purpose, that anyone would ask even half these questions of a citizen of more substantial means, or that she would consent to answering them. Her lawyer, if she had one, would point out that HIPAA, the health information privacy law, does not seem to apply. Many of the questions would not interest the actuaries and medical screeners employed by a private insurance company, and few if any doctors would require the answers to oversee a pregnant woman’s care.

Government administrators justify these extraordinary protocols by invoking the need to protect vulnerable children, a compelling state interest by any account. Medicaid and welfare information systems are, in fact, intertwined with the enforcement arm of Child Protective Services (CPS). Any sign of an unsafe environment can trigger a CPS investigation, a traumatic and even more invasive affair. No one doubts the need in some cases for intervention. The aggressiveness and tone of the screening, on the other hand, suggest a tendency to presume that poverty itself is a marker of risky parents or that poor mothers are deficient mothers until proven otherwise.

By the numbers, poor families are likelier to be investigated and likelier to be found culpable for abuse and neglect, but there is little evidence that abuse itself is more prevalent among them than in society at large. Justice Thurgood Marshall once asked, in dissent from the case that the Ninth Circuit relied upon in Sanchez: “Would the majority sanction, in the absence of probable cause, compulsory visits to all American homes for the purpose of discovering child abuse? Or is the Court prepared to hold as a matter of constitutional law that a mother, merely because she is poor, is substantially more likely to injure or exploit her child?”

Some scholars believe the psychosocial effects of poverty increase the actual incidence of abuse; others, that scarce resources make it difficult for children to handle the consequences of otherwise commonplace parental neglect; still others, that the hypervisibility of poor people—to the state—makes them likelier to draw attention from officials who are empowered or obliged to report suspected abuse.

The common denominator in each explanation is poverty.The government justifies its punitive oversight of welfare families as a vital tool for protection of poor children. But we decline to do as a society what social science identifies as the most effective intervention on behalf of child safety and health: make their families less poor. One in five American children lives in poverty; American children are ranked twenty-sixth among rich countries in terms of “overall well-being;” and we are close to the lowest spenders on benefits and services to families. The surveillance apparatus of the welfare state behaves, in effect, as a substitute for more generous aid—leaving no stone unturned in its zeal to root out bad behavior, as the system defines it, that implicitly blames the moral flaws of poor mothers for the indigence of their children.

Indigence, Compulsion, and Choice

In 1977, the Supreme Court heard the case of a young woman named “Roe” who wished to end an unplanned pregnancy. This was not the Roe who made history four years earlier in Roe v. Wade, the landmark case that established abortion as a constitutional right. In this new case, another pseudonymous woman asked the Court to strike down state barriers that left her powerless to take advantage of her namesake’s victory.

In contrast to the first Roe, who had the means to buy an abortion if only she found one for sale, the second Roe lived in deep poverty. The impact of her indigence, in court and in her prior efforts to gain control of her pregnancy, would be hard to overstate. The U.S. Supreme Court and state authorities espoused broad, universal principles of individual freedom in balance with the common good. As elsewhere in this report, the court’s abstraction was out of sync with the lived reality of the parties. Also, as elsewhere, the state made plausible claims of purpose that seemed to conceal other motives. The outcome may have forced Roe, an unwed mother of three, to deliver a fourth child against her will, and imposed a crushing financial cost on her teenage co-plaintiff.

From the moment Roe made the hard choice to terminate her pregnancy, the Connecticut Welfare Department became an implacable foe. The welfare commissioner managed federal Medicaid funds for the state, Roe’s sole source of medical support. When the U.S. Supreme Court struck down Connecticut’s ban on abortion—the nation’s oldest—the welfare agency rewrote its regulations to unsure against payment even for medical care that would otherwise qualify. Coverage of identical procedures depended on whether they were “incident to” childbirth or abortion. There was one exception, in theory, provided that Roe could secure a “certificate of medical necessity” from her doctor and a signature from the only person in the state who was authorized to approve: the chief of medical services, Division of Health Services, Department of Social Services. The record shows that Roe made the attempt, but failed.

Under directly applicable precedent, Roe had a constitutional right to privacy against state interference with the very personal choice to terminate her pregnancy. Under federal statute, she was entitled to free and comprehensive medical care. The brutality of Connecticut’s assault on those rights is best understood in concrete terms. The commissioner of welfare, responsible for the health and well-being of the state’s least powerful residents, shut down abortion among the very poor by using their indigence and his financial control to overbear their free will.

Privacy and Abortion: An Aside

Before finishing the story of Roe, we pause to address a question that many readers may now be asking themselves. What does privacy have to do with the right to abortion?

In fact, the constitutional right to privacy and its scope were central questions in both Roe cases. “Privacy,” in the sense invoked here, does not coincide with the privacy at stake in our previous discussions of policing and social benefits. In those contexts, privacy should have provided a shield against arbitrary search, seizure, surveillance, or compulsory submission to intimate questioning. For disfavored minorities and the poor, we found, the shield was so porous that it seldom kept government out.

Privacy in reference to Roe describes a more muscular right: not only freedom from unwanted eyes (and ears, and hands), but freedom to live personal lives without interference. Scholars today use the term “decisional autonomy.” That means much the same thing as the “freedom to be let alone” that Brandeis proposed as a young man, expounded in dissent as an associate justice, and died fourteen years too soon to see the Court adopt in 1965. Brandeis conceived of privacy as a boundary between state authority and sheltered domains of personal life where government did not belong. Within those domains, privacy stood for freedom to act on our choices without need for approval from society or state. We could form ill-advised friendships, read deplorable books, enjoy indecent liaisons, patronize establishments of low repute, live by unorthodox principles or none at all.

This was the privacy that the Griswold court enshrined as a constitutional right, a declaration of independence with undercurrents of “none of your business” and “don’t tread on me.” It was also, it must be said, a rich man’s vision of privacy, confident in the power of individuals to work their will on the world. For Roe, ground down by poverty, self-sufficiency could be a mirage. It was a great deal easier to form an intent than to carry it out.

Griswold introduced privacy as binding precedent, holding that government had no business peering into the marital bedroom. When the Griswold doctrine begat Roe v. Wade, women of childbearing age attained a power of decision that history had not afforded them before. With reliable access to birth control and abortion, sex and procreation became separate choices. Women gained control over life-changing events that had simply happened, ready or not, to their mothers and grandmothers. These were liberating developments in the fullest sense of the word. Women could decide when and whether to try to conceive a child, and change their minds from month to month. If a pregnancy threatened their health or well-being, they could end it.

That is, to put the point more carefully, some women could do those things.

(Poorly) Disguised Intent

The maddening fact about Roe’s plight is that anyone else in her shoes, any woman in America, would have needed only the slenderest of means to carry out her intent. A part-time paycheck, a money jar or a debit card with $300 or $400 in the bank would have transformed Roe’s life at that moment. Roe, by the case record, had no access to such resources.

The U.S. Supreme Court offered no remedy. By a vote of 6 to 3, the majority in Maher v. Roe upheld the state regulations. Edward Maher, the commissioner of social services, acknowledged that the regulations were designed to “discourage” abortion, but he justified them in the name of public health, not moral compulsion. Roe’s lawyers argued, and the court did not dispute, that the state’s financial boycott was a ban on abortion in all but name for those who had no other source of funds.

The majority opinion did not suggest much serious reflection on that point. The Court stuck to a formal distinction between direct state interference with a guaranteed right, which had not happened here, and conduct that the Court described, circuitously, as “state encouragement of alternative activity consonant with legislative policy.” By “alternative activity” the Court meant that the plaintiffs would carry their babies to term. By “encouragement” it meant, as it acknowledged, that cutting off funds could lawfully close all doors to an abortion for some women.

This appeared to leave the Court in an embarrassing bind, having acknowledged that Roe and others had been deprived of a fundamental right. It chose, therefore, to discount the burden that Connecticut inflicted on the indigent. Its understanding of cause and effect and consequence in the observable world might charitably be called incomplete. The Court reasoned, variously, that Roe’s indigence—and therefore the harm she suffered—was not the state’s fault (“neither created nor in any way affected by the Connecticut regulation”); that zeroing the budget for exactly one disfavored medical procedure does not count as “unduly burdensome”; that notwithstanding Roe’s complete financial dependence, her guaranteed right to choose “implies no limitation on State’s authority to make a value judgment favoring childbirth over abortion and to implement that judgment by the allocation of public funds”; and an unfortunate echo of Marie Antoinette in the suggestion that Connecticut’s budget choice does not diminish Roe’s alternative to seek “private abortion services.”

The Court’s allergy to constitutional analysis of the government’s spending power is understandable on its own terms. The justices may have reason to fear that constitutional review of state spending choices would bring a bottomless expansion of workload, a dearth of principled methods to draw the lines, an incursion into the proper function of the political branches, a blurring of legal and policy questions, and so on.

The fact remains, though, that the Court is left to defend the unattractive proposition that government may use the coercive power to withhold state spending from the most financially insecure citizens in order to to accomplish the same end that the state is forbidden to write into law. The Court is likewise wedded to a proposition that would not survive an introductory class on taxation or property. The Court’s reasoning leaves it no choice but to claim, as a general matter, that government has no role in the distribution of wealth and property in America. Lobbyists for the rich spend hundreds of millions of dollars to influence the language of tax rates, exemptions, carry overs and subsidies; the allocation of scarce rights on public lands and within electronic spectrums; the laws of inheritance; the finer points of trade policy; the permitting of construction and renovation and zoning; and too many other examples to name. Government does not exert sole control over any part of our economic life, but its unique powers of compulsion and its monopoly on the supply of money, among many other policy instruments, allow it to play as large or small a role as it chooses.

As for the line-drawing problem, it is probably not as hard in every case as the justices may fear, least of all in the one we speak of here. Connecticut wrote a budget of zero for exactly one, disfavored, politically controversial medical procedure in the close aftermath of a landmark decision that gave Fourteenth Amendment privacy protection to personal choice about exactly the same procedure.

Expansion of the Franchise

The success of Connecticut’s gambit helped inspire a similar, but stricter effort to starve abortion of resources, this time from Congress. The Hyde Amendment, named for principal sponsor Rep. Henry Hyde, remains in effect today. It prohibits the use of federal Medicaid funds for abortions unless the pregnancy threatens the life of the mother. Medical necessity or threats to her own health do not suffice. The Court upheld the law’s constitutionality in Harris v. McRae, ruling again that withdrawal of funding placed “no governmental obstacle in the path of a woman who chooses to terminate her pregnancy.” The Court repeated its prior assertion that the new law did not burden a woman’s privacy interest. Her poverty did that, and a destitute woman had no “constitutional entitlement to the financial resources to avail herself of the full range of protected choices.”

Taken together, the Court’s holdings in the Connecticut case and on the Hyde Amendment drew the first brushstrokes of a new “poverty exception” to the law of procreative privacy, parallel to the one we described in our treatment of policing and distribution of social benefits.

The state and federal laws, blessed by the high court, leave intact the freedom to choose an abortion in affluent communities, but compel other women to carry unwanted—or dangerous—pregnancies to term. Justice Brennan took up this argument in a dissent that, if present politics are any indication, is many years away from carrying the Court. The Hyde Amendment, Brennan wrote, was a “transparent attempt by the Legislative Branch to impose” the moral and social preferences of the majority on “a sensitive and iintimate decision that the Constitution entrusts to the individual.” More troubling, he wrote, “the Hyde Amendment does not foist that majoritarian viewpoint with equal measure upon everyone in our Nation, rich and poor alike; rather, it imposes that viewpoint only upon that segment of our society which, because of its position of political powerlessness, is least able to defend its privacy rights from the encroachments of state-mandated morality.”

At last count, three-fifths of women covered by Medicaid and aged fifteen to forty-four live in states where Medicaid covers abortion only in cases of life-endangerment, rape, or incest. Among the fifty states, only seventeen provide Medicaid coverage for abortions in cases of medical necessity that falls short of life-threatening risk; seven fully fund elective abortions that are known as “nontherapeutic” procedures. Another way to say the same thing is that only seven states give practical support for poor women to enjoy the same constitutional right to decisional autonomy that is guaranteed to more affluent ones.

Establishing a Framework for Reform

Laws and norms that safeguard privacy in the United States do not work for poor people, racial minorities, or otherwise disfavored groups. As practiced in contemporary America, surveillance leaves them open to arbitrary and encompassing intrusions that are seldom if ever inflicted upon society at large.

Why this is the case—the question of intent—is beyond the scope of this report. The systems we have examined—public safety and social welfare—are created and supported by many well-intentioned people in good faith. They need not be malicious by design, however, to have harmful effects. Cybernetic theorist Stafford Beer’s maxim that “the purpose of a system is what it does” is instructive in this regard. It’s not sufficient to dismiss the significant harms of our policing and welfare systems as unfortunate side-effects. For one thing, a model that blames every harm on righteous accident cannot plausibly cover all the abuses we have documented here. Some of of the harshest surveillance measures are features, not bugs, in the system. Discriminatory surveillance and control are not epiphenomenal to the project of providing for the poor; they are purposeful aspects of its design. And their persistence is a political choice.

Addressing the disparate harms of surveillance means developing a more capacious concept of “privacy” that is attentive to the experience of those who live beyond its practical reach.

Some of our finest thinkers on surveillance have cabined themselves off from facts on the ground in unfamiliar neighborhoods.

Some of our finest thinkers on surveillance have cabined themselves off from facts on the ground in unfamiliar neighborhoods. Inattention to the experiences of non-white, non-male, and non-rich people produces embarrassing results, even in otherwise rigorous work. In a critique of Daniel Solove’s influential Taxonomy of Privacy, for example, Ann Bartow faults him for reducing the harms of surveillance to a feeling of “discomfort” and “unease.” A similar failure of imagination is suggested by James P. Nehf’s assertions that “privacy is seldom a matter of life and death” and that “most of the injuries caused by the misuse of data in modern society are not particularly embarrassing or emotionally disturbing.” As we have tried to demonstrate, arbitrary government intrusions in the terrains of East New York or Chicago’s West Side are routinely accompanied by humiliation, physical force, the risk of a lethal encounter, or crushing financial pressure.

Traditional thinking about privacy rights has not addressed the practical limits of its reach. The trouble with the right to privacy as the catchall concept is that it imagines the bearer of such a right “as a self-determining, unencumbered individual, a being connected to others only by choice.” The gentleman in that picture bears no resemblance to the more typical targets of police and welfare surveillance that we have encountered in these pages. The presumption of self-sufficiency in the asserted right “to be let alone” is not germane to men and women whose lives are defined by poverty and interdependence. “Man’s Castle,” without a castle, is no shelter.

When privacy is conceived as an abstract, egocentric right, it becomes difficult to defend against invocations of necessity to protect important interests of society at large. As Gilliom notes, “new surveillance initiatives are almost always in pursuit of broad social goods like reducing crime, drug use, or welfare fraud,” against which “the lonely and selfish claim to privacy just can’t stand up.” Those interests are overused to justify surveillance, but they are real.

Likewise, the ostensible tradeoffs of privacy and security are frequently misconceived. Not for nothing does the Fourth Amendment begin with the words, “The right of the people to be secure….” Deprivations of privacy and security tend to walk in lockstep among poor and marginalized populations under surveillance. As we have seen, policing can instill terror; surveillance can ruin lives. When we speak of privacy and security, we would do well to ask: privacy for whom, and security against whom?

These are questions and problems that the judicial branch is unsuited to resolve. The U.S. Supreme Court’s forthcoming decision in Carpenter may or may not mark a break from decades of sterile Fourth Amendment analysis, much of it based on risible presumptions of fact about the privacy we expect and the privacy we “voluntarily” choose to waive. In broader perspective, the Court will not and arguably should not become the engine of fundamental reform in this space. Within any foreseeable future, the Court’s majority will not discover a constitutional right to privacy that is robust enough to address the abuses we have described here.

Some legal scholars hold out hope for progress by way of a doctrine against what are known as “unconstitutional conditions.” The doctrine states, in short, that government may not require anyone to cede a constitutional right in exchange for a public benefit. That might be a powerful weapon of reform if the Court were not committed to sidestepping its implications. As we have seen, the Court distinguishes fundamentally between “direct state interference with a protected activity” and a “mere refusal to subsidize a protected activity” (emphasis added). The Court acknowledged in Goldberg v. Kelly (1970) that “forces not within the control of the poor contribute to their poverty.” It accepted in Harris v. McRae that indigence creates an obstacle, sometimes even insuperable, to the assertion of certain rights. The majority of justices, however, firmly decline to take the next logical step. There is no constitutional restraint against stripping a citizen of the financial means to exercise a guaranteed right. (Recall that Mayer v. Roe the state wrote a new regulation to remove covered medical care if provided in connection with an abortion.) Medicaid and welfare spending rules can close the door to abortion in poor communities as tightly as if the procedure were a crime. The Court interprets that financial leverage as a neutral subsidy choice, implicating no constitutional right. As the majority reiterated in DeShaney v. Winnebago (1989), “the Due Process Clauses generally confer no affirmative right to governmental aid, even where such aid may be necessary to secure life, liberty, or property interests of which the government itself may not deprive the individual.”

We are not opposed to carefully chosen litigation. We root for the good guys and rely on them to think carefully about the risks of an adverse decision. We predict, even so, that politics and policy will have to carry the main burdens of reform. That is as should be. Americans do not define our social compact by the barest minimum compliance with the Constitution. We express our values and priorities—our purpose, Beers might say—in acts of choice. The federal government, for all its solicitude for the rich, still offers support to the disabled, the elderly, and the unemployed alongside its gifts to real estate developers and hedge fund managers. A majority of voting-age Americans hold protection of the defenseless as a core value. That sense of justice, so innate that we display it in early childhood, can be marshaled in the cause of surveillance reform.

There is room in a just society, as Dorothy Roberts has put it, for “not only the negative proscription against government coercion, but also the affirmative duty of government to protect the individual’s personhood from degradation and to facilitate the processes of choice and self-determination.” Defining appropriate boundaries here is no small task, but a right to privacy that covers the poor and disfavored will need to explore the terrain.

One important precondition of reform, in our view, is an intersectional approach to analysis, goal-setting, and policy instruments. This is not an empty slogan. Defining a universal right, if we want it to mean something practical, relies on a clear understanding of the particulars. Lawmakers and policy advocates are likelier to achieve the results they intend if they ask how any given proposal would interact with the real-world conditions of the rich and poor alike, man as well as woman, white as well as black, Christian as well as Muslim. Absent explicit attention to the disparate impact of surveillance tools, the liberal conception of general privacy not only leaves the disfavored behind, but also is likelier to fail on its own terms. The Fourth Amendment, to put the point lightly, already does a fine job of protecting white suburban men from stop-and-frisk and klieg lights aimed into their homes. We will have to think harder about how to bring that security to the urban poor.

Initiatives such as the Color of Surveillance delegation, the Stop LAPD Spying Coalition, and the Movement for Black Lives (M4BL) are already engaged in a version of intersectional politics. The M4BL platform, which covers a far broader landscape, includes a demand for an end to the mass surveillance of black communities, and the end to the use of technologies that criminalize and target their communities (including IMSI catchers, drones, body cameras, and predictive policing software). The platform calls for reforms at the federal, state and local levels to address the specific harms of those tools and practices. Similarly, the Stop LAPD Spying Coalition describes itself as “committed to promoting and defending the human rights of multiple, interconnected groups of people, including but not limited to immigrants, youth, African Americans, Latinas, Muslims, LGBT communities, Indigenous and First Nation People, and all oppressed and marginalized communities.”

Some reformers seek support from the privileged by portraying the unprivileged as test subjects for oppressive surveillance that may one day expand into the population at large. The argument is a well-intended effort to build a common sense of risk, but we believe it is a mistake. While it is true that aggressive new forms of are sometimes normalized over time—drones built for warfare overseas were repurposed to photograph migrants on the border, then used overhead at anti-Trump protests—too much is made of technology creep. The worst of the abuses we describe in this report are vanishingly unlikely to spill into whiter, richer, more privileged precincts where citizens have the power to fight back. There is no sense pretending otherwise, and no need: gross abuse of state power demands our attention whether or not it reaches us personally.

There is no sense pretending otherwise, and no need: gross abuse of state power demands our attention whether or not it reaches us personally.

An intersectional approach begins by observing the tangible effects of surveillance on the lives of real people. What disquiets us as we look ahead is the strong potential of new technologies, networked and integrated with the ones we have, to compound the harms of unequal treatment with new burdens on those citizens least capable of bearing them.

Consider the potential impact of “Big Data,” a term that is more often invoked than explained. Big Data is shorthand for the marriage of sophisticated new computational techniques with inexpensive storage and off-the-shelf hardware that is faster and more capable than multi-million-dollar supercomputers from just a few years ago. Big Data offers even a mid-sized company the realistic prospect of acquiring and manipulating data sets on a national scale. For the U.S. government, which can buy just about any commercial data, collect a great deal more on its own, and then share large chunks of it with state and local authorities, there is no obvious limit. Artificial intelligence and machine learning spot patterns in data sets that no human could discern in a lifetime’s work. These capabilities offer great potential for good. They also provide the building blocks of the most consequential new surveillance machine since the smartphone.

In application, Big Data techniques have already begun to amplify the disparate harms of surveillance in disfavored neighborhoods. Algorithmic analysis of historical crime data has begun a trend toward “predictive policing,” an aggressive focus on hypothetical future crime that existed only in science fiction until recently. Pouring law enforcement resources into historically over-policed communities, where the once-fashionable “broken windows” philosophy led to thousands of arrests for minor infractions, allows past injustice to define present policy. In Chicago, police have also mined Facebook and other social networks to identify individuals they judge to be at high risk of committing or becoming victims of gun violence. Even if these correlations sometimes prove out, the concentration of law enforcement power in “at risk” communities will reinforce the poisonous effects of historic biases. It aims to preempt bad acts by a small minority by subjecting all residents to aggressive law enforcement intrusions on the basis of their associations, demographics, and geography. Policy makers should give careful thought to that punitive impact on the blameless majority, which is exactly akin to the central grievance of the colonists against King George. As noted above, New York’s experiment with stop and frisk discovered firearms in 1.5 percent of its searches, many or most never used in a crime, at the cost of millions of searches that made up the other 98.5 percent. There is considerable risk that machine learning techniques, when applied to a statistical record of unequal policing, will reproduce that bias in the guise of neutral science. Prejudice embedded in computer code will be an exceptionally difficult question for lay policy makers to judge. Hamid Khan, the campaign coordinator of Stop LAPD Spying, has adapted an old programming aphorism to describe this risk: “It’s racism in, racism out.”

Two Guiding Principles for Further Research and Advocacy

Mitigating or preventing the harms of abusive surveillance may call for policies that are not typically considered under the rubric of privacy. Below we sketch two lines of further research and advocacy. Our purpose is not to write a comprehensive agenda. Our intent is to invite policy makers and privacy researchers into a conversation that takes disparities of power seriously in all its dimensions.

Meaningful Accountability for Racialized Policing

For three years, our country has immersed itself in long needed soul searching about policing and punishment. Black Lives Matter and allied groups have persuaded a growing number of cities and states to create more stringent measures of police accountability. Many members of the white majority public have come to understand that disproportionate police attention and violence are real in the places where black men and women live and work. For the first time in American history, a significant number of white people see this moral calamity for what it is. Similarly, the Sanctuary movement has built support for measures that protect undocumented residents from manifestly unjust applications of immigration rules by decoupling local policing from the federal deportation system. Attention to the broader equities, for example, might discourage surveillance and arrest of victims who arrive at the courthouse to testify about domestic abuse.

Here again, an intersectional approach counsels us to pay close attention to the particular and overlapping harms of law enforcement overreach in communities of color. Privacy advocates will find much to learn from the Movement for Black Lives, Million Hoodies Movement for Justice, Black Youth Project 100, and others. That connection to our work is obvious.

The broader agenda of this coalition ranges far afield of surveillance and privacy. The coalition has urgent equities of its own, some of them involving stakes of life and death. In their totality, however, the changes sought by advocates for racial justice aspire to reduce the scale and intensity of conflict with police. From a privacy advocate’s point of view, any progress there would be a powerful force for restraint of the most intrusive forms of search, arrest and questioning. We think of it as a surveillance dividend on proposals that might otherwise seem unrelated.

Given the disposition of the current federal government, the best prospects for law enforcement reform emerge at the local and state levels. There, prosecutors have wide discretion on the disposition of nonviolent and other petty crimes. Local governments have leeway in cooperation with increasingly aggressive Immigration and Customs Enforcement (ICE) authorities. Cities may likewise review with fresh eyes their use of gang databases that are spilling over with black and Latino residents without any substantial oversight of their methods and evidence. Local and state governments control their own criminal laws as well, offering them an opportunity to reconsider the enormous impact of drug possession laws on prison populations. They have an opportunity to reconsider the enormous impact of drug possession laws on prison populations. Interactions with police over suspicion of minor, non-violent crime accounts for the largest share of hostile intrusions upon the bodily autonomy of the poor.

The state of our national debate about racialized justice has likewise opened a space for conversation about the limits of policing as a catchall response to violence, mental health crises, and drug use in our cities. Many law enforcement authorities agree that they are being asked to do too much. “What police have been forced to do in this country is perform triage,” said Eugene O’Donnell, a professor at the John Jay College of Criminal Justice in New York and a former cop and prosecutor. Having failed, as a society, to mitigate the cumulative impact of severe inequality, racial bias, underfunded schools and historically high rates of substance abuse and mental illness, we have left police departments holding the bag. Police are not social workers; they are men and women empowered to resolve conflict and enforce order with violence if they judge it necessary. They have neither the training nor resources to perform the roles that have fallen upon them by default. People with guns should not be our first responders, nor often the last, to the symptoms of poverty, mental illness, or addiction. Any substantial improvement along these lines would radically reduce the burdens of unequal surveillance. Without progress on the former, the latter will be a lot harder to address.

Building a Welfare System That Respects the Humanity of Its Beneficiaries

In her forthcoming book Automating Inequality: How High-Tech Tools Profile, Police, and Punish the Poor, Virginia Eubanks proposes to replace America’s means-tested welfare system with a guaranteed universal basic income. The idea of an unconditional stipend, akin to Social Security and Medicare, remains far out of step with the present climate in Washington, although it does enjoy a surprising diversity of political support. It offers a useful perspective, however, on a principal source of state surveillance power. Suppose, by way of thought experiment, that every American had a guaranteed threshold of income, sufficient to support the bare necessities of food, clothing, shelter and medical care. Would something along those lines make a difference for privacy? Surely it would. Even a very modest foundation of financial security would strip welfare functionaries of the power to force their way into a beneficiary’s personal relationships and secrets. If welfare were more like agricultural price supports, that is, social benefits in the United States would resemble the ideal proposed by Eubanks: “no drug-testing, scrutiny of your parenting, or financial surveillance.” Support without conditions “assumes that poor and working-class people know best how to spend their money and care for their families.”

It is unorthodox, to put it mildly, to speak of economic and social changes of this magnitude in the name of privacy. It is hard to deny, on the other hand, that our national experience with means-tested support is incompatible with a meaningful right to privacy and human dignity. We are unable to imagine a guide for caseworkers that could teach them to respect personal boundaries while burrowing into the financial, romantic, and medical secrets of applicants. The authors of this paper are not of one mind about the merits of a universal stipend writ large, but we are entirely agreed on this: it would remove at one stroke what is arguably the federal government’s most unjustly burdensome form of domestic surveillance.

It is hard to deny, on the other hand, that our national experience with means-tested support is incompatible with a meaningful right to privacy and human dignity.

A universal child allowance might be an effective interim step, on its own terms and as a window into the surveillance effects of welfare. The Century Foundation’s Rediscovering Government Initiative has produced essential research on the beneficial impact of this allowance on child poverty. In the privacy sphere, any unconditional entitlement would break down walls that divide a stigmatized “beneficiary class” from the rest of society, and therefore chip away at support for overweening surveillance in that community.

The strongest connection of privacy and financial support was suggested by the story, told earlier in this report, of the pseudonymous Roe fought and lost her abortion case in the U.S. Supreme Court. The state used its financial power to thwart her exercise of a constitutionally guaranteed choice, and the Court went along. An intersectional approach to privacy would consider how that narrative might have changed if Roe had not been obliged to rely on the same hostile bureaucracy to feed her child and provide basic medical care.

Conclusion

We are indebted in this report to many more thinkers and advocates than we have cited or have room to name. Many of them have been working at the intersections of privacy, inequality, poverty, race, and surveillance for decades. These voices, many of them named in our footnotes, deserve a wider hearing in our public discourse about surveillance reform.

If nothing else, we intend this report as a provocation to our fellow privacy and civil liberties advocates. We ask you to consider the implications for your work, empirical and conceptual, of the grossly disparate burdens of surveillance across lines of race and wealth and power. When you think about the scope of state power, have you taken sufficient account of its routine abuse? The authors believe as strongly as ever in privacy as a foundational requirement of self-government and personal freedom. What we do not believe is that a universal approach to surveillance policy is sufficient to reach its goals.

Source: https://tcf.org/content/report/disparate-impact-surveillance/