by Lawrence Winnerman

On May 28, 2021, the United States Senate voted on whether to investigate the mob that had hunted its own members through its own hallways four months earlier.

Fifty-four senators voted yes. Thirty-five voted no. Six Republicans—Murkowski, Collins, Cassidy, Romney, Sasse, Portman—crossed over to stand with every voting Democrat, four months after they had all crouched together behind the same doors.

The measure failed.

In the United States Senate, 54 beats 35 and loses. And notice what did not happen that afternoon: Nobody rose to speak against the commission. Nobody held the floor through the night, voice cracking, aides ferrying water. The senators who killed the investigation of the attack on their own workplace were not even required to be in the building. The blockade was a piece of paper—a cloture motion that failed to find 60 votes—and the filibuster, the word we still use for it, was an email that arithmetic sent on somebody’s behalf.

Stack up what waits behind that number. The Voting Rights Act restored: needs 60. Statehood for the four million Americans with no Senate voice: needs 60. Automatic registration, real donor disclosure, public financing, the anti-corruption package that polls like free ice cream: 60, 60, 60, 60. Half of the 43 reforms in this catalog—the entire structural middle of this series—stack up behind a single procedural gate.

The gate is not in the Constitution. It is not in any statute. No court has jurisdiction over it and no state ever ratified it. It is a Senate rule, and 51 senators can change it on any Tuesday they find the nerve—a fact both parties have already proven, twice, this century.

The easiest mechanism in American politics guards the hardest gate. And it was built by accident.

The Accident

Start with what the framers actually believed about supermajorities, because they told us, repeatedly, in writing.

They had just lived under the Articles of Confederation, which required nine states of 13 for anything important, and the country had nearly died of it. Hamilton, in Federalist 22, wrote that supermajority requirements serve “to embarrass the administration” and “destroy the energy of the government.” Madison, in Federalist 58, went further: If more than a majority were required to legislate, “the fundamental principle of free government would be reversed”—power transferred to the minority. So they wrote exactly five supermajorities into the Constitution—treaties, veto overrides, expulsions, impeachment convictions, and amendments—and the precision of that list is the point. Name the exceptions, and everything else is majority rule.

The Senate legislated that way for its first generation. Then, as he left office in 1805, Aaron Burr—tidying the chamber’s rulebook on his way out—recommended deleting a motion he considered redundant clutter: the previous-question motion, the tool for ending debate and forcing a vote. The next year, in 1806, the Senate obliged. Nobody noticed for three decades that the chamber had thrown away its off switch. The first true filibuster arrived in 1837, and the Senate spent the next 111 years unable to close debate at all.

The most consequential act of housekeeping in American history was a deletion.

The patch came in 1917, after a dozen senators talked President Wilson’s armed-ship bill to death on the eve of war and Wilson raged about “a little group of willful men” rendering the government helpless. Rule 22: debate could be closed—by a two-thirds vote. In 1975 the threshold dropped to three-fifths. Sixty.

That is the entire constitutional pedigree of the number that now governs what may become law in America: a housekeeping error, a wartime patch, and a haircut.

Subscribe now

The Weapon

A tool is what it does. So look at what the filibuster did across its first century and a half, once senators understood what the missing off switch made possible.

In 1922, the Dyer anti-lynching bill passed the House, had the votes in the Senate, had a president ready to sign—and died to a filibuster. So did the anti-lynching bills of the 1930s. So did the poll-tax bans, five of them. So did the fair-employment bills. Historians who have counted find that between the end of Reconstruction and 1964, the bills the filibuster killed outright were, almost without exception, civil rights bills. Not tariffs, not spending, not war powers—those all passed or failed by majority like the framers intended. The supermajority gate had a specialty, and the specialty was keeping Black Americans out of American democracy. Strom Thurmond still holds the record: 24 hours and 18 minutes, in 1957, against a voting rights bill.

In 1964 the segregationists held the Senate floor for 60 working days—the longest filibuster in the chamber’s history, against the Civil Rights Act. On June 10, 1964, the Senate invoked cloture, 71–29. It was the first time in the rule’s 47-year existence that cloture had ever been invoked on a civil rights bill. The rule had not changed. What changed was outside the building: Birmingham, the March on Washington, a movement that made the cost of inaction higher than the cost of action. Column Zero, in one roll call.

The saucer where the Senate supposedly cools its tea was, for four decades, the place anti-lynching bills went to die.

Share

The Setting

Through all of that, at least, obstruction cost something.

A filibuster was a body on the floor, a voice giving out, a caucus visibly strangling a bill in public while the rest of the nation’s business waited. The cost was the check: You could count the great filibusters of the twentieth century on your fingers because mounting one meant paying for it.

Then, in the early 1970s, the Senate renovated—with the best intentions, which is how most load-bearing walls get pulled. To keep obstructed bills from freezing the whole calendar, Majority Leader Mike Mansfield institutionalized a two-track system: Set the filibustered bill aside, move on to other business, come back later. Humane. Efficient. Catastrophic. Because if the Senate simply moves on, nobody has to hold the floor. Nobody has to speak. The filibuster stopped being something a senator did and became something a bill had—a property, a setting, a flag in the database. The talking filibuster did not die by vote; it died of convenience.

The numbers tell it plainly. For decades in the mid-twentieth century the Senate saw a handful of cloture votes per Congress—entire two-year terms with one or two. The current Congress has seen 268 cloture motions filed, and it is not done. That is not 268 impassioned stands of conscience. That is a default. The Senate’s operating assumption is no longer majority rule with exceptions; it is 60, with exceptions—and every bill that cannot find a supermajority dies silently, unmarked, without a fingerprint on it.

Nobody has to stand anymore. Nobody even has to speak.

Leave a comment

The Locks Already Opened

Now the part that gives the game away: the exceptions.

Budget reconciliation, created in 1974, moves taxes and spending at 51 votes—which is why every signature economic law of the last decade, from the 2017 tax cuts to the 2021 rescue plan to the 2022 climate bill, passed with a simple majority and no apology. In November 2013, a Democratic majority used a point of order and 51 votes to end the filibuster for executive appointments and every judge below the Supreme Court. In April 2017, a Republican majority used the identical move to finish the job for Supreme Court seats. Trade deals move at a majority. Congressional Review Act resolutions move at a majority. War powers resolutions move at a majority.

The carve-outs, laid side by side: judges, justices, taxes, budgets, nominations, trade. Everything the modern Senate actually prioritizes already passes at 51. What remains behind the 60-vote wall is ordinary legislation—laws, the general-purpose product the institution exists to manufacture. The Senate has quietly concluded that majority rule is safe enough for lifetime appointments to the Supreme Court but too reckless for a voting rights bill. Stated that way, no one would defend it. So it is never stated that way.

And every carve-out used the same mechanism, the one sitting in the drawer right now: a point of order, a ruling of the chair, 51 votes to sustain the new precedent. No amendment. No court. No governor, no legislature, no ratification clock.

The Senate is the sole owner of its rules, and it has renovated them, this century, whenever a majority wanted something badly enough.

Subscribe now

Some Will Say

Some will say: the Senate is the saucer that cools the coffee. The saucer story is a secondhand anecdote, and even taken at face value it describes the Senate itself—six-year terms, staggered classes, statewide constituencies, the natural brake of a smaller chamber. It says nothing about a 60-vote threshold invented in 1917 and lowered in 1975. The Senate cooled plenty of coffee for 130 years without a routine supermajority. Deliberation is a schedule. The filibuster is a veto.

Some will say: it protects the minority. The minority keeps its debate time, its amendments, its committees, the courts, the states, and the next election—every legitimate instrument of opposition. What it loses is a costless, invisible veto. And the arithmetic of whom the veto serves is not subtle: 41 senators from the 21 smallest states represent 11% of the American population. A device that lets 11% of the country overrule the rest is not minority rights. It is minority rule with a procedural alibi.

Some will say: you will miss it when you lose. Honestly: yes. Majorities I despise will pass laws I hate, and I will hate them at full volume. That is called losing an election, and its remedy is winning one—not rigging the machine so that nobody can govern at all. Notice, too, which way the current flows. A party whose agenda is tax cuts and judges already gets everything at 51; reconciliation and the nomination carve-outs saw to that. The filibuster’s surviving jurisdiction—voting rights, labor, civil rights, climate—is not a neutral list, and it never has been. One side loses far more to this rule than the other, and it is the side that keeps defending it.

Some will say: without it, nothing forces compromise. The record is in, and it says the opposite. The 60-vote Senate does not produce compromise; it produces nothing—and nothing is the minority’s win condition, because the majority owns the failure at the polls. A majority that must deliver or answer for it has every reason to negotiate. A minority that profits from gridlock has none.

The threshold does not moderate outcomes. It moderates accountability.

Share

The Repair

The fix is not exotic, and it is not even abolition.

End the 60-vote threshold for final passage. Keep debate—and restore its price: a talking filibuster requiring continuous, physical presence on the floor, capped at some civilized outer bound, two weeks, after which the Senate votes and the majority carries. End the silent hold, the anonymous single-senator veto that survives on pure custom. If a minority believes a bill is a catastrophe, let it stand up, on camera, on the record, and say so for as long as its convictions and its knees can bear. Conviction was never the problem. The discount was.

Mechanically, this is the smallest reform in the entire catalog. It requires no court, no state, no amendment, no cooperation from anyone outside one caucus of one chamber. The precedent-setting move has been executed twice in living memory, once by each party, and took an afternoon both times. Everything else on this ladder needs something—ratification, litigation, trifectas, decades.

This needs 51 people and a Tuesday.

Leave a comment

The Alibi

Which leaves the real question—not whether the filibuster should go, but why it has not.

The polling is soft, the history is ugly, the framers are on record, and the carve-outs have already conceded the principle. Something else is holding the door.

The something else is that the filibuster is the Senate’s alibi. It lets a senator campaign on the whole catalog of promises and deliver none of it, gesturing at the number on the way to the fundraiser. It spares members votes they would rather not cast and fights they would rather not have. It converts every hard question into a procedural inevitability that is somehow no one’s fault. Both caucuses have found this useful. Both caucuses lean on the door. The lock stays on the gate not because anyone loves it but because everyone, on the worst days, is grateful for the excuse.

Which is why this rung, the easiest mechanism on the whole ladder, still obeys the first law of this series: The mechanism is never the scarce resource. In 1964, cloture was 47 years old and had never once been invoked for civil rights—and then Birmingham and the March and Freedom Summer made 71 senators find a nerve the chamber had misplaced for half a century. The rule did not change. The pressure did. A majority breaks this lock on the day it fears its voters more than it loves its alibi, and not one day sooner.

Behind the gate: the Voting Rights Act. Two new stars on the flag. Your registration, protected. The money, named. Bodily autonomy. Nearly every rung above this one leans, one way or another, on this one.

Last week I said to bring a torch. Keep it. You will not need it for the lock—the lock opens from inside, 51 hands on the bar, and it always has.

The torch is for finding out who is holding the door.

« Rung 1: The Winnable Seven | Rung 2: The Master Lock | Rung 3: Register Everyone, Purge No-One »

MORE FROM LAWRENCE WINNERMAN

by Lawrence Winnerman

On February 13, 2020, the chairman of the Senate Intelligence Committee sold stock in 33 separate transactions in a single day—somewhere between $628,000 and $1.7 million of his holdings, including hotel chains. His committee was receiving daily classified briefings on a novel coronavirus spreading out of Wuhan. One week later the market began a slide that erased a third of its value.

He said he had relied solely on public news reports. The Justice Department investigated and brought no charges. The Securities and Exchange Commission probed and let it go. The Senate Ethics Committee said nothing at all.

Nothing happened, because under the rules, nothing was supposed to.

Americans noticed. Banning members of Congress from trading individual stocks polls at 86%—87% among Republicans, 88% among Democrats, numbers that agree with each other more than the parties agree on anything else. It is, by a comfortable margin, the most popular idea in the entire 43-reform catalog this series is built on.

So here is a thing that happened this January, six years after those 33 trades. The House Administration Committee finally advanced a stock-trading ban. It moved on a party-line vote, over the minority’s objection, and when you read it you find the mechanism: Members must stop buying new stock—but they may keep every share they already own, and sell whenever they like with a week’s public notice. A competing, stricter bill sits behind a discharge petition dozens of signatures short. The floor vote leadership promised for the first quarter came and went with the quarter.

The fight was never about whether to ban congressional stock trading. Four in five Americans settled that argument years ago. The fight is happening where these fights always happen—in the fine print, in a definitional clause, in a room with no cameras.

Welcome to Tier I. Seven reforms sit on this rung of the ladder, and they share three properties. Each needs only a statute or a chamber rule—no amendment, no court, no interstate compact. Each is either wildly popular or so boring that no one would march against it. And each has already died at least once—never at the vote, always in the fine print. Seven reforms, seven scorecards, taken at speed—because the pattern matters more than any single repair: The ribbon-cutting is never the fight.

1. Ban the Trading, for Real

Congress already banned insider trading by its members once. The STOCK Act of 2012 passed in a wave of public fury, and it worked the way a screen door works on a submarine: It required disclosure of trades within 45 days, punishable by a late fee of $200, routinely waived. Members kept trading. Some got faster.

The real version has one non-negotiable word in it, and the word is divestiture. Members, spouses, and dependent children sell their individual stocks—into index funds, Treasuries, or a blind trust that is actually blind, meaning the assets are sold and replaced, not parked behind a curtain the owner already knows the shape of. That last clause is where the January bill quietly surrendered: a “ban” that grandfathers existing holdings is a press release with a penalty schedule. The portfolio a member built over 20 years in office is the conflict of interest. Letting them keep it is not a compromise; it is the whole game.

Subscribe now

2. Give Congress Its Brain Back

In 1995, the incoming House majority zeroed out the Office of Technology Assessment—Congress’s own in-house research arm, roughly 140 staffers whose entire job was telling members what was true about complicated things. Its budget was about $22 million, which at today’s rate of federal spending is roughly one hundred seconds. The authorizing statute was never repealed. The office is still legally alive, sitting there in the U.S. Code, waiting for an appropriations line.

Meanwhile the median congressional staffer leaves the Hill inside three years, usually for the lobbying firm across the street, which pays triple. By one political scientist’s arithmetic, corporations now spend more lobbying Congress than taxpayers spend funding it. A legislature whose institutional memory turns over every three years cannot out-argue the people writing its bills for it—and that is not a metaphor. Lobbyists draft legislative text every week in this country. Somebody has to.

This is the cheapest, most upstream repair in the catalog: fund the OTA, raise staff pay to parity, rebuild the committee expertise that used to exist. It is a precondition for every flashier reform in this series, not a competitor with any of them. And nobody will ever march for it, which is exactly why it is worth writing down.

Share

3. Make Them Read the Bill

In December 2022, the text of a $1.7 trillion omnibus spending bill—more than 4,000 pages—was released in the small hours of a Tuesday morning. The Senate passed it that same week. No member of Congress read it. That is not cynicism; it is arithmetic. And into bills like that one, at 2 a.m., go the riders nobody voted for and nobody can be blamed for, because everybody was voting for the whole Christmas tree or shutting down the government.

The repair is procedural and old-fashioned: single-subject rules, 72 hours of public posting before any vote, restored open markups—the boring regular order that both parties praise in the minority and abandon in the majority. But the catalog’s stress test found where this one dies, and it is not in the rulebook. Every one of these rules can be waived, and the waiver belongs to leadership. A rule the leadership can waive at 2 a.m. is not a rule. It is a suggestion with a press release. Reform that does not move waiver power out of leadership’s hands—to a supermajority of the chamber, to anyone—is cosmetic.

Leave a comment

4. Fix the Courthouse Plumbing

There is a single-judge division in Amarillo, Texas, where filing a civil case is not a lottery. It is an appointment. For years, every case filed there landed on the same district judge—which is why, when activists wanted a nationwide ruling yanking a drug off the market that the FDA had approved 23 years earlier, they did not file in Washington or Dallas. They filed in Amarillo. One judge, chosen by the plaintiffs, briefly got to govern the medicine cabinet of every American.

The fix is plumbing: random case assignment districtwide, so no litigant can pick their judge, plus the 66 new district judgeships the federal courts have been begging for across a decade of rising caseloads. A bill doing the judgeships—the JUDGES Act—passed the Senate unanimously in 2024. Unanimously. Then the House sat on it until after the presidential election, passed it only once the election had decided who would appoint the judges, and the outgoing president vetoed it at Christmas. The most bipartisan reform in this entire catalog died of timing. The lesson is transferable and the catalog applies it everywhere: stagger the judgeships across future administrations, set the start dates past the next election, and pass it in the one season nobody knows who benefits.

5. Put Teeth on Court Ethics

A sitting Supreme Court justice accepted, over two decades, luxury travel from a billionaire with interests before the Court—private jets, a superyacht, island-hopping in Indonesia—and disclosed essentially none of it until reporters did the disclosing for him. The Court’s response, eventually, was to adopt an ethics code. Its enforcement mechanism is the honor of the people it regulates. It has no other.

Congress cannot force a justice to recuse; the constitutional ice gets thin fast. But two levers sit squarely inside Congress’s power, and the catalog is blunt that they are the only two: money and sunlight. An inspector general for the judiciary housed outside the Court’s control. Real-time disclosure of gifts and finances—30 days, not never. Appropriations consequences for noncompliance. None of this touches a justice’s vote or seat. All of it touches the thing that made the flights invisible.

Get more from Lawrence Winnerman in the Substack app

Available for iOS and Android

Get the app

6. Make Every Election Provable

More than nineteen out of every twenty American votes are already cast on or backed by paper. What most states still do not do is use the paper: a risk-limiting audit—hand-checking a statistically rigorous random sample of ballots against the machine count—turns “was it stolen?” from a cable-news shouting match into a checkable claim with a confidence interval. It is the cheapest trust technology democracy has ever invented, and only a handful of states require it.

The catalog’s finding on this one is about sequencing: the binding constraint is not evidence quality but whether the losers pre-committed to the mechanism before they knew the outcome. An audit adopted in a contested December is an accusation. The same audit adopted in a sleepy off year is furniture—and furniture is the goal. You cannot retrofit legitimacy. You can only install it in advance, in a boring year, while nobody important is looking.

Give a gift subscription

7. Coup-Proof the Count

On January 2, 2021, an American president spent an hour on the phone with Georgia’s secretary of state asking him to “find 11,780 votes.” In the years since, county officials in Arizona, New Mexico, and elsewhere have discovered a simpler tool: just refuse to certify. Sit on your hands past the deadline and let the lawyers inherit the wreckage. Two county supervisors in Arizona were criminally charged for trying it in 2022. The next attempts will be better organized.

Congress patched the top of this stack in 2022—the Electoral Count Reform Act raised the objection threshold and made the vice president’s role explicitly ceremonial. But the bottom of the stack, where the votes actually become results, still runs on the assumption that county canvassing boards act in good faith. The repair kit: make certification a nondiscretionary ministerial act—a math-reporting duty, not a judgment call—with an automatic 48-hour court order when anyone refuses; criminalize intimidation of election officials; lock chain of custody to funding. And the threat model is already moving: if certification becomes ministerial, the next play is jurisdiction-shopping—redefining who certifies, midcycle, until the duty lands on someone willing to breach it.

Refer a friend

Seven Rooms

The last boxes of the seven scorecards, read together: a definitions clause. An appropriations line no one fights for. A waiver in the leader’s pocket. A calendar. A court reviewing itself. A retrofit that arrived too late. A moving target.

Seven reforms, and not one of them dies in public. Not one dies at the vote—the votes, when they happen, are landslides. They die in small rooms, in the fine print, in the gap between passing a thing and keeping it. Which is what Column Zero promised you: The ribbon-cutting is never the fight. The fight is the word “blind.” The fight is who holds the waiver. The fight is a start date.

Suppose we win all seven. Every scorecard on this rung still tops out at three out of five, because Tier I repairs the outlets and the junction boxes—real work, worth doing, worth winning. It does not touch the panel.

Everything above this rung—the Voting Rights Act, statehood, the money, the maps, nearly half the catalog—runs through a single gate, and the gate is guarded by a number: sixty. One Senate rule, changeable by 51 senators on any Tuesday they choose, stands in front of more democratic repair than every reform in this essay combined. It requires no court, no state, no amendment, no one’s permission but the Senate’s own.

The easiest mechanism in American politics, and the hardest politics in America.

Rung 2 is the master lock.

Bring a torch.

Subscribe now

« Essay 0: Power Comes First | Essay 1: The Winnable Seven | Essay 2: The Master Lock »

Coming Monday

Everything you just read—seven repairs Congress could pass at speed—rests on one assumption: that a bill can actually reach a vote. Monday we climb to the first real gate on the ladder, Rung 2: The Master Lock. It’s the single reform standing in front of half this list, and the fight over it is the fight over whether any of the rest is even possible.

And Monday, something else opens here—a room for the readers who want to help build this, not just read it. I’ll tell you everything then.

—Lawrence

MORE FROM LAWRENCE WINNERMAN

by Lawrence Winnerman

On the night of November 6, 2018, Florida voted to forgive.

Amendment 4 restored the vote to Floridians who had completed felony sentences—nearly 1.4 million people, the largest single expansion of the franchise in this country in half a century. It needed 60% to pass. It got 64.55%. In a state that decides presidencies by a single point, in a midterm, across every kind of county, close to two-thirds of voters looked at their neighbors and agreed on something simple: A debt paid is a debt paid.

Seven months later, the Florida legislature passed Senate Bill 7066, and the governor signed it. The bill declared that “all terms of sentence” included every fine, fee, and court cost ever attached to a conviction—money owed, in most cases, not to any victim but to the machinery of the court system itself, and money the state conceded in federal litigation it often could not even calculate. Roughly 774,000 of those 1.4 million people were disenfranchised again.

This time by an invoice.

The reform passed. The reform lost. And the gap between those two sentences is what this series is about.

The Series I Am Not Writing

There is a genre of democracy writing I have come to think of as the wish list. You have read it. I have read it. Ranked-choice voting, statehood, term limits, overturn Citizens United—each item worthy, each item floating free of every other, none of them weighted, none of them sequenced, as if a constitutional order were a gift registry. The wish list has two defining features: nothing in it is scored against anything else, and nothing in it explains why none of it has happened.

This series is the other thing. Over the past months I built a catalog of 43 serious reforms—everything from banning congressional stock trading to rewriting the Constitution itself—and stress-tested every one of them against the questions the wish lists skip. How much would it actually fix? How hard is the lift? How long would it take? Through what mechanism—a statute, a Senate rule, a state compact, an amendment? And the question the wish lists never think to ask, the one Amendment 4 answers with 774,000 invoices:

What breaks it?

Every essay in this series carries the same scorecard: Fix Power, Lift, Horizon, Mechanism, What Breaks It. The last box is the honest one, because the finding that kept recurring—across election law and court reform and money and maps—is that passing a reform is rarely the hard part. Keeping it, enforcing it, and stopping its capture is the war. Florida taught that lesson in seven months. The scorecard exists so we never have to learn it twice.

The essays run as a ladder, 13 rungs, from the winnable to the wild. Rung 1 is a set of bills Congress could pass this year with the members it already has. Rung 13 is the question of whether an eighteenth-century architecture can be patched at all, or only replaced. You can climb as far as your politics will carry you. I am going all the way up.

But before the ladder, this column. Column Zero. Because the strongest argument I encountered in months of stress-testing was not an argument for any reform on the list.

It was an argument against the entire frame.

Subscribe now

The Dissenter

The argument goes like this. Every mechanism on my ladder—every statute, every rules change, every compact and amendment—quietly assumes something that does not currently exist: a legislature or an electorate already willing to pass it.

The scarce resource in American reform has never been the right idea. The scarce resource is organized people willing to spend years forcing the idea through, and then more years defending it from the people it disarmed.

Check the record. The Wagner Act did not create the labor movement; the strike waves of 1934—Toledo, Minneapolis, the San Francisco waterfront—created the Wagner Act. The Civil Rights Act did not produce the movement; Birmingham produced the Civil Rights Act, and Selma produced the Voting Rights Act, and the marchers were on the bridge before the bill had a number.

The Nineteenth Amendment arrived after 70 years of organizing, not before it. Reform, every time we have actually gotten it, has been a lagging indicator of power.

No mechanism organized the flying squadrons in Minneapolis. No scorecard walked across the Edmund Pettus Bridge. No white paper ever survived contact with a legislature that had no reason to fear the people proposing it.

The dissenter is right.

And I am writing the series anyway. Here is why.

Share

Why Write the Wiring Diagram at All

This series is called Rewiring the Republic, and I mean the name literally enough to take its physics seriously. A wiring diagram does nothing on its own. Copper does not carry current because the blueprint is elegant. You can draw the cleanest schematic ever committed to paper, frame it, publish it, win awards for it—and the house stays dark until somebody generates power and pushes it through the wall.

But the reverse failure is just as real, and we have lived it more recently. In the summer of 2020, somewhere between 15 and 26 million Americans took to the streets—by most counts the largest protest movement in the country’s history. It produced city council resolutions, corporate pledges, a summer of promises.

At the federal level it produced one major bill, which died in the Senate, at 60 votes, the same place nearly everything dies. All that current, and it hit a wall nobody had mapped as immovable.

That wall is Rung 2 of this series.

So the two failure modes of American reform turn out to be mirror images. The wish list without a movement: a blueprint for a house with no power hookup. The movement without a blueprint: a million volts grounding out against a wall it did not know was structural. Power without a diagram gets spent badly. A diagram without power is just paper.

This series is the diagram drawn for the people building the current. It will tell you which walls bear load and which are drywall. It will tell you which repairs are cheap and real, which are expensive and transformative, and which beloved fixes are cosmetic. And at every rung it will point back to this column, because every “What Breaks It” box on every scorecard, traced far enough down, arrives at the same answer: Organized power showed up to defend the reform, or it did not.

Leave a comment

The Rules of the Ladder

A few promises so that you can hold me to them.

I will score honestly, even when it hurts the cause. Some reforms progressives adore have modest ceilings, and I will say so. Early voting is good, and it is not transformative. Some reforms nobody marches for are so crucial to the engineering of our democracy we cannot skip them—who certifies your county’s vote count matters more than almost anything with a hashtag—and I will say that too.

I will name the three master locks now. The filibuster, which gates half of everything in this catalog. The Supreme Court, which hangs over every statute the way weather hangs over a picnic. And single-member districts, the quiet arithmetic that makes two-party trench warfare mathematically inevitable no matter how fairly anyone draws the lines. Three of the 13 rungs belong to these locks alone, and every other essay will link back to them.

And I will not hand you villains where the problem is structure. There are people behaving badly in this story—you will meet some of them—but the reason American democracy is failing is not that bad men exist. Bad men have always existed. The reason is that the architecture now rewards them, reliably, at scale, and architecture is something you can fix.

This is not partisan cheerleading; half the reforms in this catalog were Republican ideas within living memory. It is not doomer despair either.

It is a field guide—the thing you could hand a skeptical neighbor without apologizing for it.

Subscribe now

Back to Florida

Because the Amendment 4 story does not end with the invoice.

The people who ran that campaign—returning citizens, led by a man who had once slept on the streets of Miami and later stood in the well of a courtroom as a law school graduate—did not file a lament when SB 7066 came down. They raised money, tens of millions of dollars, and started paying off court debts, one returning citizen at a time, and registering them, one returning citizen at a time.

The state changed the rules after losing the game. The organizers absorbed the hit and kept working, because organizing is the one thing in this story that sabotage cannot repeal.

That is the posture this series assumes. Not optimism. Stubbornness with a blueprint.

Rung 1 is next: seven repairs Congress could pass tomorrow—a stock-trading ban that four out of five Americans want, a brain Congress amputated in 1995 for the price of a hundred seconds of federal spending, a fix for the courthouse where one judge in Amarillo gets to govern the country. Seven bills, seven scorecards, and seven small rooms where each of them goes to die.

Every rung on this ladder is copper until someone throws the switch.

Power comes first.

« Foreword: The Guardrails Were Us | Essay 0: Power Comes First | Essay 1: The Winnable Seven »

MORE FROM LAWRENCE WINNERMAN

by Lawrence Winnerman

In the last week of January 2025, 17 inspectors general—the federal government’s internal watchdogs, the people whose entire job is catching fraud and abuse inside the agencies they oversee—were fired in a single night, most of them by a two-line email. Federal law is not ambiguous about this. It requires 30 days’ advance notice to Congress, in writing, with case-specific reasons, before an inspector general can be removed.

No notice was given. No reasons either.

Nothing happened. Some letters were sent. Some statements expressed concern. The law requiring the notice was still on the books the next morning, unrepealed, fully in force, and fully irrelevant.

It had simply stopped being important, because the truth was that there was never an enforcement mechanism in place.

What the Guardrails Were Made Of

For most of my life I believed—because I was taught, carefully and well, by people who believed it themselves—that American democracy was a machine full of guardrails. Checks and balances, presented to a kid in the 1970s as something close to physics. Watergate was the proof in my childhood: the system worked, we were told, because a president who broke the law was made to leave. It did not occur to me until embarrassingly recently that the sentence had a quieter second half. The system worked because that president’s own party decided it should. Because senators walked up the drive and told him it was over. Because he agreed to go.

The last decade has been a public stress test of every assumption in that civics lesson, and the second presidential term we are now living through has finished the demolition. One by one, the things we called guardrails have turned out to be made of something softer than we thought: norms. Customs. Gentlemen’s agreements between people who stopped being gentle some time ago. The 30-day notice. The income tax releases. The recusal. The blind trust. The precedent. The way it’s always been done. Each of them held for decades—not because anything enforced them, but because everyone agreed to behave as if they were walls.

A guardrail that depends on the driver’s goodwill is not a guardrail. It is a suggestion painted on the road.

The guardrails were us. They were only ever us.

That sentence can land two ways, and the entire posture of this series depends on which one you choose. It can land as despair: Nothing was ever protecting us. Or it can land as assignment: We were the enforcement element all along, and that responsibility has not gone anywhere. I choose the second reading, and not out of temperament. Out of history.

Subscribe now

The House We Inherited

The Constitution of 1787 is a house—a genuinely brilliant house, designed by people who had just escaped a landlord and thought hard about doors and locks. And its wiring was state of the art for the eighteenth century: information moving at the speed of a horse, power flowing through institutions designed to be slow on purpose, because slowness was the safety feature.

We still live in that house. We have added rooms—27 amendments’ worth. We have swapped fixtures by statute a thousand times. But the core wiring is original, and the load has changed. The house was wired for candles, and we are running data centers on it. Decisions that the founders assumed would take a season now happen in an afternoon; the president’s hand rests on switches the framers never imagined one hand would reach; information—true and false—moves through the walls at a speed no breaker in the original design can trip. I have written elsewhere in this publication about what happens when democratic institutions meet machine speed. This series is about the wiring itself.

And the maintenance is overdue in a way we have stopped noticing. The last constitutional amendment conceived and ratified by living Americans was the Twenty-Sixth, in 1971. (The Twenty-Seventh, ratified in 1992, was a leftover proposed in 1789—a curiosity, not a repair.) If you are younger than 73, the Constitution has never been freshly amended in your adult lifetime. We treat that as normal. It is not normal. It is deferred maintenance on a 239-year-old electrical system, and deferred maintenance does not stay deferred. It compounds. The shorted circuit takes the wall with it.

Share

It Has Been Done Before

Now the part the doom headlines never include.

Americans have rewired this house before—repeatedly, structurally, and almost never in calm weather. The Progressive Era put four amendments into the Constitution in eight years, including two that fundamentally redistributed power: the direct election of senators and the income tax. Women won the vote by amendment in 1920 after 70 years of organizing. And between 1961 and 1971—one decade—Americans amended the Constitution four separate times: presidential votes for the District of Columbia, the abolition of the poll tax, presidential succession, and the vote for 18-year-olds. That last one, the Twenty-Sixth, went from congressional proposal to full ratification in about a hundred days. The fastest constitutional surgery in American history was performed in 1971.

The country that did that: Vietnam at full boil. Cities that had burned two summers running. A president the other party loathed. Assassinations still fresh in the national memory, a generation in open revolt against its parents, and a political establishment certain the republic was coming apart. Those Americans—frightened, furious, divided in ways we would recognize instantly—looked at the house and rewired it anyway. Four times.

They were not better than us. They were not calmer than us. They were just earlier.

And now the honest other half, because this series does not deal in pep talks. It is also true that the amendment path is harder today than it has ever been. Ratification takes 38 states, and we have sorted ourselves into a country where 13 legislatures can block anything, and reliably would. The parties have nationalized; the muscle has atrophied from 50 years of disuse; and the same fraught, contentious moment that makes the repairs urgent makes the biggest tools hardest to reach. Both of those things are true at once—it has been done before, and it may be harder now than ever.

This series lives in the space between those two sentences.

Leave a comment

Not Optional

Which brings me to the quiet lie underneath most democracy writing: the word “optional.”

Reform as a preference, an enthusiasm, a genre for people who listen to podcasts about the Federalist Papers. Wouldn’t it be nice. Someday, maybe. And for a long time the lie was affordable, because the norms were holding and the house, however creaky, kept the lights on.

The norms are not holding. We have now watched—are watching, in real time, this year—what the house does under a tenant who probes every wall for what is actually load-bearing and what was only ever paint.

The answer, too often, has been paint.

That knowledge does not expire when any one tenant leaves. Every future occupant of every office now knows exactly which restraints were real. The 17 inspectors general are not an anecdote; they are the new baseline. What one president does without consequence, the next president starts from.

So no—rewiring is not optional, and it is not partisan, any more than a fire code is partisan. Some of these repairs are cheap and could pass this year. Some require rebuilding the service panel. A few require asking whether the house needs a second founding, and I will take even those seriously when we get there. But the choice on the table is not reform versus the status quo. There is no status quo. The wiring is failing at the pace of the loads we put on it. The choice is repair, or keep discovering—one shorted circuit at a time, one ignored statute at a time—what was never actually attached to anything.

Here is how we will do it. The series is built as a ladder—every repair American democracy needs, bundled and ordered from easiest to hardest. It starts with bills Congress could pass this fall with the members it already has. It ends at the hardest questions a constitutional democracy can ask itself. Every essay scores its reforms the same way, on the questions that actually matter: how much it fixes, how hard the lift, how long it takes, and what breaks it. You do not have to believe in the top of the ladder to climb the bottom of it. You just have to start climbing.

Subscribe now

The Why

I am 56 years old. My generation got the tail end of the confident civics—the Bicentennial, the filmstrips, Schoolhouse Rock, the machine that fixes itself—and has spent its adulthood watching the machine not fix itself. We were told the wiring would outlast us. It will not, not without work, and I have stopped waiting for the electricians of legend to show up. There are no electricians coming. There is us, a stack of blueprints, and a house that is worth it.

The moment is precarious. The guardrails were us. The repairs are not optional. And people more frightened than we were did harder repairs than these, in worse weather than this.

The ones who rewired this house before us were not braver than we are.

They were just earlier.

It’s our turn at the panel.

MORE FROM LAWRENCE WINNERMAN