Every argument we have about voting in this country is an argument about the electorate. Who’s on the rolls. Who got purged. Who needs an ID, who needs a stamp, who needs a signature match, who gets a Sunday and who gets a Tuesday. We fight about the size and shape of the voting public with the confidence of people who believe the whole of self-government is contained in that one question.
The men who wrote the Constitution would have found this bewildering. Not because they thought the franchise unimportant, since they argued about it constantly and in the states they argued about it viciously, but because for them it was a subordinate question. It sat downstream of a prior one:
What is government permitted to do to a person, no matter who votes for it?
That’s the question the American constitutional order is organized around. Everything else, the House and the Senate and the electors and the courts and the states and the amendment process, is machinery built in service of it. The Founders built a system that could survive the will of the people. Discovering that will was the easy half of the problem, and they spent far less time on it.
That distinction is the entire point, and we’ve almost completely lost it, which is why our politics now consists of two coalitions each convinced that the answer to being governed badly is to win the next election by more than the other guy.
The Founders had a different diagnosis. They thought the problem was the sword itself, and that who happened to be holding it mattered a good deal less than we assume.
What a right is
Start where they started, because if you get this wrong nothing downstream makes sense.
The American founding rests on a claim so unfashionable now that it has to be stated slowly. Rights exist before the state does. No charter creates them, no legislature confers them, no vote generates them, and the state’s legitimacy is measured by whether it respects them.
The Declaration says this in language so familiar we don’t hear it anymore. Men are endowed with rights, endowed by a source outside politics. Governments are instituted to secure those rights. The rights come first and the government is the instrument. Reverse that order, and you’ve written a different country, the kind our Founders overthrew.
Blackstone, whose Commentaries sat on every serious American lawyer’s shelf, opens Book One with a chapter titled “Of the Absolute Rights of Individuals,” and reduces them to three: the right of personal security, the right of personal liberty, and the right of private property. Absolute meaning inherent, belonging to the person as a person and not to the subject as a subject. Parliament could protect them or violate them. Parliament could not manufacture them.
Locke put the sharper point on it. A government is a trust, and the people who hold power hold it the way a trustee holds an estate: for a purpose, subject to terms, revocable when the terms are broken. The legislature, Locke insisted, can’t be arbitrary even when it’s popular. It can’t rule by extemporary arbitrary decrees. It can’t take a man’s property without his consent. Those aren’t restrictions a legislature has generously accepted. They’re the conditions of its existence, and a legislature that breaks them has surrendered its right to exist.
Here’s why any of this matters to an argument about voting. If rights precede government, no vote count can reach them. A right that fifty-one percent can dissolve is a permission that hasn’t been revoked yet, by the state or by the mob. Every serious constitutional argument in American history is a variation on that sentence.
What a vote is for
So what’s the vote, if it isn’t the source of our rights or government power?
It’s a defensive instrument, and that’s a smaller answer than we’re used to hearing.
The vote does three things in the Founders’ scheme. It supplies consent, so the people who bear the burdens of law have some hand in making it. It supplies accountability, so officials who abuse power can be removed without bloodshed. And it supplies information, so grievances reach the government before they reach the streets.
Notice what’s missing from that list. Nothing on it turns a mob preference into justice. A majority that votes to seize your house hasn’t made the seizure right, only likely, and it’s made itself illegitimate in the process. The Founders understood the ballot the way a sensible man understands a lock. It is indispensable against what comes through the door, and no help at all with the wiring in your own walls.
There’s a second thing about voting that the whole classical tradition insisted on and we’ve almost entirely forgotten. Voting is an exercise of power over other people.
When you choose your religion, your work, your wife, your reading, you’re exercising liberty, disposing of what’s yours. When you vote, you help dispose of what’s everyone’s. You’re participating in the coercive apparatus that will tax your neighbor, conscript his son, license his trade, and imprison him if he refuses. That’s a different kind of act; it carries a different weight, and it generates a different set of obligations.
This is the premise underneath every restrictionist argument in Western political thought, and, properly understood, underneath every argument for civic education too. If the ballot were just private preference, competence would be irrelevant, because nobody demands qualifications for choosing a hobby. It’s precisely because the vote reaches other people that anyone ever thought to ask who should hold one.
The wheel
Before the Founders wrote anything, they read. And what they read told them that popular government has a characteristic way of dying.
Aristotle gave them the vocabulary. Six regimes in the Politics, three sound and three corrupt, paired off. Kingship corrupts into tyranny. Aristocracy corrupts into oligarchy. And politeia, the balanced constitutional rule of the many, corrupts into what Aristotle called democracy, meaning rule by the many in the interest of the poor instead of the common interest. Read that vocabulary carefully, because it sits upside down from ours. In Aristotle’s scheme, democracy is the diseased version of the healthy form.
Polybius supplied the motion. His account in Book Six describes a cycle. Monarchy decays into tyranny, is overthrown by the best men into aristocracy, decays into oligarchy, is overthrown by the people into democracy, and democracy decays into mob rule, which exhausts itself until some strongman restores order and the wheel turns again. Rome’s achievement, on his reading, was jamming the wheel. Three elements, monarchic and aristocratic and popular, mixed into one constitution so that each restrained the others. His own description is that the balance was struck so finely nobody could say for certain whether the thing was an aristocracy, a democracy, or a despotism.
One caution before somebody corrects you with it. Polybius uses the words differently than Aristotle does. For him democracy is the healthy form and its corruption is ochlocracy, mob rule. What both men share is the structure underneath the vocabulary: the rule of the many comes in a sound version and a rotten one, and the rotten one arrives by a predictable route.
Montesquieu turned this into a true political method. Power has to be arranged so that power checks power, and his statement of the danger is worth having in the original: constant experience shows us that every man invested with power is apt to abuse it at some point, and to carry his authority as far as it will go. Not evil men. Every man. Which means the design problem was never selecting good officials. It was guaranteeing that bad ones couldn’t do much damage.
The Founders inherited all of this and drew the obvious conclusion. Tyranny has three possible addresses, because rule has only three possible configurations: the one, the few, and the many. A monarch can crush you. An aristocracy can crush you. A majority can crush you. The mechanism differs, and either way you still get crushed.
We’ve been trained to fear the first two and to treat the third as a contradiction in terms. The Founders expected the third to be the one that showed up in America, because the Revolution had already excluded the other two.
The case for competence
Now to the arguments themselves, starting with the oldest and, in some ways, the most uncomfortable.
If the vote is an exercise of power over others, then the ignorant exercise of it harms others. John Stuart Mill made this case about as forcefully as anyone ever has. In Considerations on Representative Government he argued that the suffrage is a trust and not a right, that the voter acts as a kind of public officer, and that a public officer’s incapacity is everybody’s problem. His own formulation beats any paraphrase of it. A man’s vote, he wrote, “has no more to do with his personal wishes than the verdict of a juryman.”
Mill’s proposal was plural voting. Everyone gets a vote and the educated get extra ones. He wasn’t trying to exclude the working man; he wanted the franchise extended, including to women, which in 1861 was not a safe position to hold. What he wanted was to keep the sheer numerical weight of any one class from becoming irresistible. A broad electorate with a counterweight inside it.
The steelman is real. An electorate that can’t identify its own government’s structure, that votes on slogans it can’t define, that’s systematically misinformed about the basic facts of the policies it endorses, is not going to preserve anybody’s liberty, its own included. Anthony Downs gave the phenomenon its name in 1957: rational ignorance. Since one vote almost never decides an election, the individual voter has no material incentive to invest in understanding anything. Calling that a moral failing misses what’s producing it. It’s a predictable response to the incentives.
And the rebuttal is just as real, and it’s fatal. Every competence test ever devised has been captured by the people administering it. That’s structural rather than accidental, and no amount of better drafting fixes it. Whoever writes the test decides the electorate, and whoever decides the electorate has already won. The American record here isn’t ambiguous. Literacy tests and understanding clauses were sold as competence measures and used as racial instruments, applied by registrars with total discretion and no appeal, and in Williams v. Mississippi in 1898 the Supreme Court blessed the arrangement on the reasoning that the tests applied to all voters equally. Facial neutrality on the page, discretion doing the work at the desk. The tool did exactly what a tool with no check on it will always do.
I should say the part that cuts against my own witness. Mill would have required voters to read, write, and do arithmetic, and he’d have disqualified non-taxpayers and undischarged bankrupts besides. The man who made the strongest philosophical case for competence proposed a literacy test. That isn’t a gotcha about Mill. It’s the argument’s own failure mode showing up in the life of the person who stated it best.
Which is, of course, precisely the Founders’ point. A power with no check on it gets abused, including a power designed to improve the electorate. Mill’s argument survives as a case for civic education. It dies as a case for gatekeeping.
The case for independence
The property qualification is usually caricatured, and that’s a shame, because the real argument is more interesting than “rich men should rule.”
The eighteenth-century case rested on independence. James Harrington argued in The Commonwealth of Oceana in 1656 that political power follows the distribution of property, and that you can’t build a stable republic on an economic structure that contradicts it. The republicans who followed him drew the practical inference. A man who owns his own land, his own shop, his own tools, votes as he judges best. A man wholly dependent on a landlord, an employer, or the Crown votes as his dependency dictates, whatever he privately believes.
The concern was never that the poor were morally inferior. It was that the dependent aren’t free agents, and a ballot cast under obligation is the patron voting twice. The secret ballot, when it finally arrived, was an attempt to solve that same problem by a different route.
Mill extended the logic somewhere more provocative. He objected to those receiving parish relief (welfare) voting, on the grounds that they’d be voting to dispose of funds contributed by others. The claim there is about the structure of the transaction, not about anybody’s character: when the beneficiary of a transfer also authorizes the transfer, the ordinary check on spending disappears.
The rebuttal, again, is decisive. Property has never reliably produced independence, since the wealthy are as capable of dependency on state favor as anyone and rather better positioned to arrange it. Corporate welfare extended a potential $1.8 trillion in energy subsidies under Biden, for example. What’s more, property qualifications in practice entrenched whoever already held property, which is the classical definition of an aristocracy. Property owners could simply change the law to prevent any new voters. And the argument proves far too much: extend it consistently and you disenfranchise every contractor, every subsidy recipient, every pensioner, and eventually everybody who benefits from any public expenditure, which is all of us even if the subsidy is just public services.
Hold onto the underlying observation anyway, because it doesn’t go away when the qualification does. The Founders’ worry was that dependency corrupts the vote. Universal suffrage relocates that worry without answering it.
The case against the coalition
Which brings us to the argument that’s aged best, the one that doesn’t depend on any claim about anyone’s character, and that classical liberals have been making with mounting evidence for two centuries.
Suppose everybody votes. Suppose the electorate is educated, independent, and public-spirited. The structural problem survives all of that: in a system where the majority can vote itself transfers from the minority, politics stops being about protecting rights and becomes a competition over the distribution of other people’s property.
Tocqueville saw the shape of it before there was much of it to see. In the chapter he titled “What Sort of Despotism Democratic Nations Have to Fear,” he described an immense and tutelary power that doesn’t oppress so much as it manages, that keeps men in what he called perpetual childhood, and that the citizens themselves choose because it relieves them of the trouble of self-government. Nobody has to stage that. People pay for it. His word was tutelary, which beats the “soft despotism” everyone says now, because a tutor is somebody you hire.
The twentieth century supplied the mechanics. Mancur Olson worked out the asymmetry in The Logic of Collective Action: a concentrated group with a large per-head stake organizes easily, and a dispersed group with a small per-head cost doesn’t organize at all. Three hundred firms fighting for a tariff will lobby ferociously. Three hundred million consumers paying nine dollars more won’t notice. The concentrated interest wins even when it’s a minority and even when the total cost exceeds the total benefit. Run that across every industry and program, and you get a state that grows continuously without anybody ever deciding to grow it. While we’re all looking for a villain, the mechanics keep on churning out big government on their own.
Hayek’s contribution was separating two things we habitually confuse: law and legislation. Law in the older sense is general, abstract rules applying to everyone alike, discovered and not invented, telling you what you may not do to your neighbor. Legislation is particular, specific commands directing specific resources to specific parties. A parliament that spends its days on legislation has stopped making law. And Hayek said plainly that his objection was never to democracy, only to unlimited democracy, an assembly with a majority and no boundary, which he thought dissolved self-government instead of fulfilling it.
James Buchanan and Gordon Tullock gave it the cleanest frame in The Calculus of Consent, and it’s the frame the Founders would have recognized instantly. There are two levels of political choice. There’s choice within rules, the ordinary business of elections and legislation. And there’s choice of rules, the constitutional level, where you decide what the game will be. At the constitutional level you don’t know where you’ll end up, whether you’ll be in the majority or the minority, taxed or subsidized. That uncertainty makes people reason more fairly, so the rules get set at that level, made hard to change, and then left alone while the ordinary game runs underneath them.
That’s the American Constitution described in the language of twentieth-century economics. The Founders got there a hundred and seventy years early, and they got there from Polybius.
The case for the burden
One more strand, briefer but persistent. Some in the republican tradition held that the franchise should track demonstrated civic commitment, on the theory that men who bear public burdens govern more prudently. Militia service. Taxpaying. Jury duty. Local office. Attendance at the town meeting.
The intuition is that skin in the game disciplines judgment, and that a citizen who has stood a watch or served on a jury or paid a tax he could feel understands in his body what he’s authorizing when he votes.
As a franchise restriction, this collapses for the same reason the others do, because whoever defines the qualifying burden defines the electorate. Can someone not physically fit to serve, due to disability, not be permitted to vote? As a description of what makes self-government work, it’s close to unanswerable, and it points somewhere the restrictionists never went. The remedy isn’t withholding the ballot from men who lack civic experience. It’s rebuilding the institutions that used to generate civic experience: the local boards, the juries, the associations, the volunteer companies, the churches and lodges and mutual-aid societies where Americans once learned self-government by practicing it on a scale small enough to fail safely.
That’s the same conclusion the Tocqueville argument reaches from the other direction. Competence is a culture, and cultures get built from the bottom.
Elections without liberty
The strongest empirical evidence for the Founders’ position comes from the last fifty years.
Fareed Zakaria named the thing in Foreign Affairs in 1997, and V-Dem now publishes an electoral democracy index and a liberal democracy index as two separate numbers, because they measure two different things and the world keeps producing places where one is high and the other isn’t. States hold genuine elections with real turnout and real ballots, and also jail journalists, capture courts, harass opposition parties, rewrite constitutions on a majority’s schedule, and hollow out the civil service. The elections aren’t always fraudulent. Sometimes the majority genuinely wants all of it.
That’s the clean refutation of the idea that suffrage and liberty are the same thing. If they were the same thing, illiberal democracy would be a contradiction in terms. It turns out to be one of the more common forms of government on earth.
Run it in reverse, and it’s just as instructive. There have been regimes with narrow franchises and unusually secure private liberty, and regimes with universal franchises and none at all. That isn’t an argument for authoritarianism; it’s an argument against the idea that voting is the apex of human liberty. The size of the electorate correlates with the security of a person’s rights far more weakly than we assume, and far more weakly than independent courts and enforceable limits do.
Which is what the Founders said would happen. Elections are necessary. They were never sufficient. Anyone who tells you the vote is the guarantee of your freedom is describing a country that doesn’t exist.
The arguments that fail their own test
A great many historical objections to universal suffrage weren’t the arguments above. They were exclusions based on race, on sex, on religion, on inherited rank, and they were often defended in the borrowed vocabulary of competence and independence. That borrowing is why the philosophical arguments are radioactive today, and the association isn’t unfair. The literacy test was a real instrument of a real regime, and the men who administered it knew exactly what they were doing.
But here’s the thing worth being precise about, and it’s a strong defense of the Founders’ framework, not a concession against it. Those exclusions fail by the Founders’ own premises.
If rights are natural, if they inhere in persons as persons and are endowed by a source outside politics, then they can’t depend on ancestry or sex, because ancestry and sex are facts about bodies and you can’t build a principle out of a fact about a body. An exclusion grounded in a characteristic instead of a civic argument contradicts the Declaration’s premise directly, which is a considerably bigger problem than offending modern sensibilities. It says the endowment is selective, which is to say it isn’t an endowment at all but a grant, which puts you back in the world of charters and permissions the Revolution was fought to leave.
And this is precisely how the expansions were argued. Frederick Douglass didn’t ask America to abandon the Founders’ principles. On July 5th, 1852, in Corinthian Hall in Rochester, he simply demanded that it apply them, holding the Declaration up against the auction block and letting the contradiction do the work. The women at Seneca Falls in 1848 didn’t draft a new philosophy. They rewrote the Declaration of Independence, lifting the preamble nearly verbatim, changing four words in the famous sentence so that all men and women are created equal, and then running their own list of grievances against a government that had refused them. The abolitionist and suffragist arguments were the founding’s own premises collecting a debt. This is perhaps the most beautiful thing about America’s founding: that even in its failures, it provided the future the tools necessary for correction.
The competence argument and the independence argument are genuine philosophical positions with genuine philosophical rebuttals. The exclusion arguments were never in that category. They fail the test the Founders themselves supplied.
Structure, not selection
So what was the Founders’ answer? If you can’t secure liberty by curating the electorate, and you can’t secure it by expanding the electorate, what’s left?
Architecture. And the architecture is astonishing once you look at it as a single machine instead of a list of civics-class facts.
They divided power by function. Legislative, executive, judicial, separated because Madison held in Federalist 47 that the accumulation of all three in the same hands, “whether of one, a few, or many, and whether hereditary, self-appointed, or elective, may justly be pronounced the very definition of tyranny.” Read the middle of that sentence again. Or elective. An elected assembly holding all three is as much a tyranny as a king holding all three, and Madison wrote the exception out before anyone could ask for it.
They divided power by level. The federal government received enumerated powers and the rest stayed with the states and the people. Two governments, each able to obstruct the other, neither able to reach the citizen from every direction at once.
They divided the legislature against itself. Two chambers, different constituencies, different terms, each capable of stopping the other. Federalist 63 is explicit that part of the Senate’s purpose was to check the people’s own temporary passions, to suspend “the blow meditated by the people against themselves” until reason could catch up. Moments when a popular majority would later be grateful somebody had made it wait.
And here’s the piece almost everyone misses. They divided power across time. House members serve two years, presidents four, senators six, judges for life on good behavior. Four institutions, four clocks, four selection methods, four constituencies. A faction that captures the country in one election captures one clock. To capture the government it has to sustain itself across a decade and more, by which time, Madison bet, it will have cooled, splintered, or been outlasted. That inconvenience to power is by design.
They put certain things beyond the reach of any of it. No bills of attainder, no ex post facto laws, the writ of habeas corpus protected against suspension, all three sitting in Article One before the Bill of Rights was drafted. Then the Bill of Rights itself, phrased as prohibitions on government and not as grants to citizens. Congress shall make no law. Shall not be infringed. Shall not be violated. And the Ninth and Tenth Amendments, which exist to say that the enumeration isn’t exhaustive and the silence isn’t a grant.
And they made the rules hard to change. Article Five requires supermajorities twice over, two-thirds to propose and three-fourths to ratify. They built the amendment power precisely because they knew they were fallible. The supermajorities are there because the rules of the game shouldn’t be alterable by whoever happens to be winning at the moment. That’s Buchanan and Tullock’s two-level insight, written in 1787.
Every one of those is a limit on what majorities can do. None of them says a word about who gets to vote.
What we did to it
A design is a set of load-bearing walls, and walls come down when people don’t know what they’re holding up.
The four clocks have been partially synchronized. The Seventeenth Amendment moved the Senate to direct popular election in 1913, and it passed for real reasons. State legislatures deadlocked over Senate seats, with Delaware going 217 ballots across 114 days in 1895 and then sitting without a senator for two years, and bribery was at least as large a driver. Those were genuine problems, and the amendment genuinely addressed them. But understand what got traded. The Senate had been built to have a different constituency, not merely a longer term. It represented states as states, whole bodies, which put an institutional defender of federalism inside the federal government. After 1913, both chambers answer to the same population on different schedules.
The separation by function has been folded back together. Consider a modern administrative agency. It promulgates rules carrying the force of law. It investigates and prosecutes violations of those rules. And it adjudicates the resulting cases before its own officers, applying its own interpretation. Legislative, executive, judicial, one institution. The people staffing it are largely competent public servants doing technical work a part-time legislature can’t do, which is exactly why it happened. But Madison’s warning was about the arrangement. He said nothing about the character of the officials inside it, and he wrote no exception for accumulations staffed by experts with good intentions.
And the enumeration has been stretched until it barely binds. The logic of Article One, Section Eight is that a list which doesn’t exhaust the possibilities is a limit. After Wickard and Helvering and two centuries of expansive readings of the commerce and spending powers, it’s hard to name much Congress couldn’t reach if it wanted to badly enough. The Court has pushed back at the margins since 1995, in Lopez and Morrison and in the part of Sebelius holding that the commerce power can’t compel a purchase, and none of that restored the original limit. A list that includes almost everything is no list at all.
None of this was a coup. Every step was taken by elected officials (most of whom were reelected), ratified by voters, upheld by appointed judges, and welcomed by majorities who wanted what the step delivered. Which is precisely the mechanism this essay has been describing.
The walls didn’t fall. We removed them. For what we thought were good reasons. We did it on purpose, by voting, and we reelected the offending politicians.
The third tyrant
Set the whole thing out and the Founders’ position resolves into something coherent, uncomfortable, and considerably more radical than the version taught in school.
They believed the vote was necessary, insufficient, and dangerous in proportion to how much power it could reach. Sacred never entered into it. And they didn’t fear kings especially. Remember, they had just defeated the largest empire on the planet, along with its king. No, they feared unchecked power, and they understood that unchecked power is indifferent to the number of hands holding it. One man, a few men, or a hundred million men. The question is never how many. The question is always whether anything stops them.
Jefferson put it in a line Madison thought important enough to quote at length in Federalist 48, and the sentence just before the famous one is the harder of the two. It will be no alleviation, Jefferson wrote, that these powers will be exercised by a plurality of hands and not by a single one. A hundred and seventy-three despots would surely be as oppressive as one.
Then the line everybody knows. An elective despotism was not the government we fought for.
Read it again. He isn’t warning about a coup, or a foreign power, or a restored monarchy. He’s warning about a legislature the people chose, doing whatever it likes, with a mandate from those people. He’s warning about us.
That’s the third tyrant. Not the one on the throne, not the few in the drawing room. The many at the ballot box, arriving with clean hands and a certified vote count, and our entire political vocabulary is now constructed to make him unnameable.
So when we fight about voting, and we will keep fighting about voting, the Founders’ question isn’t who’s on the rolls. Their question is the one we’ve stopped asking.
What does the winner actually get?
If the winner gets everything, every institution and every clock and every check, the whole machine at once, then it doesn’t matter who votes. You’ve built a throne and hired a search committee. The wheel Polybius described will turn on schedule, and it won’t care how many people were consulted.
If the winner gets a term, a chamber, a portion, bounded by rules he can’t rewrite and courts he doesn’t own, then the electorate can be as broad as human dignity requires, because no majority, however large, can reach the thing that matters.
We talk about the Constitution as a machine for counting. It was built as a machine for stopping.
Everything since has been an argument about whether we still want one.