On the morning of June 6, 2013, readers of the Washington Post and the Guardian opened their browsers to find that the National Security Agency had been secretly collecting the phone records of tens of millions of Americans, not suspected criminals, not foreign nationals, not people under investigation, but ordinary American citizens whose only connection to any investigation was that they used a phone on a network operated by Verizon.
The legal authority for this collection: a secret order from the Foreign Intelligence Surveillance Court, a judicial body whose proceedings are classified, whose opinions are sealed from the public, and whose judges are appointed without Senate confirmation. The statutory authority: Section 215 of the PATRIOT Act, which allows collection of records “relevant” to a terrorism investigation, a relevance standard that the secret court interpreted to encompass essentially the entire Verizon customer database.
The program had been running for years. The legal opinions justifying it were classified. The members of Congress who had been briefed on it were prohibited from discussing it publicly. The Americans whose records were being collected did not know.
This is not a story about a rogue agency or a single administration’s overreach. It is a story about the Federal Farmer’s distinction between rights on paper and rights in practice, and about what happens to civil liberties when the government possesses the power, the technology, and the legal architecture to work around the rights it is formally obligated to honor.
The Warning That Rights on Paper Are Not Rights in Practice
The Anti-Federalists demanded a Bill of Rights as a condition of ratification. But several of the most careful writers were explicit that they were not confident a bill of rights would actually protect liberty. They demanded it anyway, as a gesture of principle, as a floor below which the government could not legally sink, but they warned that rights enumerated in a document were only as strong as the structural constraints on the government that was obligated to honor them.
The Federal Farmer was the most direct on this point. Writing in October 1787, before the Bill of Rights existed, he argued that rights protected only by text, without structural enforcement mechanisms, without independent courts with real power to check the executive, without a government small and close enough to the citizenry to be meaningfully accountable, would become aspirational rather than operational. The government would find ways to work around them.
He wrote:
“A bill of rights is of real use in establishing fixed principles… but the practice of arbitrary men and of arbitrary governments may be very different… It is one thing to have a right secured on paper, another to have it secured in practice.”
He was not predicting the NSA. He was predicting the dynamic: a government with vast powers, distant from ordinary citizens, facing genuine security pressures and bureaucratic incentives toward expansion, would interpret written rights as narrowly as it could justify and its capabilities would consistently run ahead of its formal constraints.
The Fourth Amendment, ratified in 1791, reads:
The right of the people to be secure in their persons, houses, papers, and effects, against unreasonable searches and seizures, shall not be violated, and no Warrants shall issue, but upon probable cause, supported by Oath or affirmation, and particularly describing the place to be searched, and the persons or things to be seized.
Particularly describing the place to be searched. The persons or things to be seized.
The FISA Court order collecting Verizon’s records did not particularly describe a place to be searched. It described a database. It did not name a person to be surveilled. It named a corporation and required that corporation to produce all records for all customers over a period of months. This is not what the Fourth Amendment’s text describes as a constitutional search. The government’s legal argument was that it was, and that argument was made in a classified proceeding that the people whose records were collected could not observe, contest, or appeal.
How Secret Law Works
The most significant civil liberties development of the past two decades is not the specific programs that have been revealed. It is the construction of a body of secret law, legal interpretations, issued by classified courts and executive branch lawyers, that govern the rights of citizens but cannot be read by those citizens.
The FISA Court was created in 1978 to provide judicial oversight of foreign intelligence surveillance. Its original design: a specialized court that would review government requests for surveillance warrants targeting foreign agents on American soil. Warrants would be issued on probable cause, reviewed by judges with clearances, classified to protect sources and methods.
What developed over the following decades was something different: a court that issued not just individual warrants but broad programmatic opinions, interpretations of law that governed entire surveillance programs affecting millions of people, written in classified opinions that the public could not read and that members of Congress had limited ability to access or discuss.
Senator Ron Wyden, a member of the Senate Intelligence Committee, spent years warning publicly that there was a gap between the public understanding of surveillance law and the secret law the government was actually operating under. He could not say what the gap was because the information was classified. He was in the position of knowing that citizens were operating under a set of rules they had never seen and could not challenge.
This is the Federal Farmer’s distinction made concrete: the Fourth Amendment on paper says one thing. The legal architecture that has developed around it, classified court opinions, executive branch memoranda, statutory interpretations that bear little resemblance to the ordinary meaning of the text, says something quite different. The citizen lives under the second body of law while being told, formally, that the first one protects them.
The Technology Problem
The Fourth Amendment was written for a world of physical searches: officers entering premises, examining papers, seizing physical objects. The warrant requirement was designed to prevent the general warrants that the colonial government had used to conduct sweeping searches of homes and businesses without specific justification, the specific grievance that the Fourth Amendment was designed to address.
The modern surveillance state does not require entering premises. It does not require physical seizure of papers. It requires a request to a telecommunications company, which has already collected the records as a byproduct of providing service, and the provision of those records in digital form to a government database where they can be analyzed, searched, and retained indefinitely.
The Supreme Court has spent decades working out how the Fourth Amendment applies to technologies the founders could not have imagined. The doctrinal history is complicated and still unresolved. What is clear is that the technology has consistently run ahead of the legal framework, the surveillance capabilities that the NSA, FBI, and other agencies could deploy by 2013 were not constrained by a legal framework designed in 1978 for a narrower problem, and the legal framework that developed to fill the gap was written in classified proceedings that the people affected could not review.
The Federal Farmer did not predict surveillance technology. He predicted the dynamic: a government with the capability to do something, facing institutional incentives to do it, will find the legal justification. Rights on paper are a constraint; they are not an absolute bar when the government controls the legal architecture that interprets what the rights mean.
Both Parties, One Direction
The surveillance apparatus that was revealed in 2013 was built across administrations of both parties and authorized by legislation passed by congressional majorities of both parties.
The PATRIOT Act was passed in October 2001 with 98 votes in the Senate and 357 votes in the House. It was signed by President Bush and reauthorized by Congress repeatedly, including under President Obama. Section 702 of the FISA Amendments Act, which authorizes collection of foreign communications that incidentally capture American citizens’ communications, was passed in 2008 and has been reauthorized by bipartisan majorities in 2012, 2018, and 2024.
The pattern after each disclosure of surveillance overreach follows a consistent script: congressional outrage, executive branch defense, reform legislation that imposes some limitations while reauthorizing the underlying authority, and continuation of the programs in modified form. The architecture, secret courts, classified legal opinions, mass collection with after-the-fact review, remains intact.
Republicans have been critical of surveillance programs when they believed those programs were used against political allies. Democrats have been critical of surveillance programs when they revealed overreach against civil liberties organizations. Both parties have voted to reauthorize the underlying authority. The institutional momentum in the national security agencies runs in one direction: toward more capability, broader authority, and more comprehensive collection. The political incentive for both parties, when in power, runs in the same direction: the legal responsibility for a security failure is easier to manage than the political responsibility for restricting the tools that might have prevented it.
What Rights in Practice Would Require
The Federal Farmer’s framework suggests the question to ask: not what the Fourth Amendment says on paper, but what structural conditions would make it operative in practice.
A right against unreasonable search is only a right in practice if there is a meaningful mechanism to enforce it, a court that can hear the challenge, a plaintiff who has standing to bring it, a remedy that provides real relief when the right is violated.
The current architecture defeats each of these. Citizens who have had their records collected under classified surveillance programs typically do not know they were collected, because the programs are classified and the government is not required to notify them. Without knowing they were surveilled, they cannot challenge the surveillance. Courts have largely held that plaintiffs who cannot prove they were specifically surveilled lack standing to challenge programs, even when the programs are publicly known to affect millions of people. The state secrets doctrine allows the government to terminate civil litigation before discovery when the case would require revealing classified information.
The result: the Fourth Amendment right against unreasonable search exists on paper and is nearly unenforceable in practice against the national security apparatus, not because of any single decision but because of the accumulated legal architecture that surrounds it.
What rights in practice would require: elimination of secret law governing citizens’ rights, any judicial interpretation that determines the scope of constitutional protections should be public, even if specific operational details of surveillance programs are classified. Warrant requirements for domestic surveillance that reflect the Fourth Amendment’s text rather than narrow interpretations developed in closed proceedings. Civil remedies that are actually accessible when rights are violated, not pre-empted by doctrines that prevent the case from reaching merits. And oversight mechanisms with genuine teeth, not briefings to cleared committee members who cannot discuss what they learn, but oversight that produces real accountability when programs exceed their authorization.
None of these exist in sufficient form. The architecture has been built over decades, and it serves the institutional interests of the agencies that operate within it too well to be easily dismantled by any administration that will itself rely on those agencies.
The Quiet Surveillance
The disclosure in 2013 produced a significant public response. There were congressional hearings, editorial debates, diplomatic consequences, and several years of genuine reform conversation. Section 215 bulk phone records collection was eventually curtailed by the USA FREEDOM Act in 2015.
And then, largely, the conversation moved on.
The surveillance architecture remains in place. Section 702 was reauthorized in 2024 with expanded authorities. The FISA Court continues operating with classified proceedings. The FBI’s access to Section 702-collected data for domestic investigations, one of the most contested provisions in the reform debates, has been authorized and renewed.
The ordinary American who sends emails, makes calls, and uses the internet does so within a surveillance environment whose scope they cannot know, whose legal justification they cannot read, and whose operation they cannot contest. The Fourth Amendment, on paper, says this requires a warrant. The law that actually governs it says something classified.
The Federal Farmer told us to expect exactly this gap. He was not warning about technology he could not have imagined. He was warning about the nature of governments with vast powers, distant from the citizenry, operating in domains where the ordinary mechanisms of democratic accountability do not function.
He asked for structural constraints that would make rights real rather than nominal. He was overruled in favor of a Bill of Rights that said the right things on paper and left the enforcement to a structure he did not trust to honor it.
He was right to be skeptical.
Federal Farmer, Letters II and IV, October 9 and October 12, 1787. Brutus, Letter II, November 1, 1787. Available at the Avalon Project, Yale Law School, and the Library of Congress American Memory collection.