Somebody is going to say we tried this.
They will be right, and it is worth hearing the objection in its strongest form before answering it.
Forty-three state constitutions contain a single-subject rule. Forty-one of them apply it to all legislation. Most date to the nineteenth century. They were adopted to restrain logrolling, expose riders, give legislators and citizens notice of what a bill actually contained, and stop the accumulation of privately tailored law.
They have produced, in the words of the leading survey of the field, literally thousands of court decisions and no consistent doctrine at all.
Legislatures evaded them. Courts enforced them unpredictably. Omnibus bills survived. Riders survived. Logrolling survived. Sometimes a court struck legislation years after enactment. Sometimes a court approved a combination broad enough to make the restriction look decorative. One political scientist describes the rule in Minnesota as effectively dead for the last quarter century.
All of that is accurate history, and none of it describes the system proposed in this series.
The record is not forty-three experiments with line-addressed law. It is forty-three variations on a much weaker instruction:
Put only one subject in each bill.
That instruction never defined the legal unit. It never required one operative proposition per line. It never required a separate vote on each proposition. It never gave a proposition a permanent address. It never built anything capable of detecting a conflict before enactment. It never connected a judicially discovered defect to a mandatory legislative repair.
The states did not build the machine and watch it fail.
They painted a warning label on the old machine and waited.
What the Nineteenth Century Was Actually Fighting
The rule has a specific birthday and a specific enemy.
New Jersey wrote the first single-subject clause into its 1844 constitution. Ohio followed in 1851, adopting the one-subject rule as one of several new checks on legislative power alongside the governor’s veto. By the turn of the century thirty-six states had joined. The count reached forty-three.
The problem they were solving was not abstract. State legislatures of that era ran on special legislation — charters, franchises, land grants, exemptions written for a single named beneficiary — and on the practice of stapling those favors to bills nobody could vote against. Railroad money moved through statehouses in exactly the way lobbying money moves through Washington now, and the drafters of those constitutions wrote the rule because they had watched it happen.
They diagnosed the disease correctly.
The California Supreme Court restated that diagnosis in 1987 with more precision than most modern commentary manages. The clause exists, the court said, to prevent legislators from combining several proposals into one bill so that by combining their votes they produce a majority for something that would have failed if the proposals had been separated.
That is the exact mechanism this series was written to attack, described accurately by a state supreme court forty years ago.
And the rule does still bite occasionally. Missouri’s Supreme Court recently struck down a bill that reduced county financial-disclosure requirements and carried, tucked inside it, restrictions on state anti-homelessness funding and a provision making it a misdemeanor to sleep or camp on state land without authorization. County accounting reform on the label. Criminalizing homelessness in the box.
The rule caught that one. It should have. A legislature that wants to criminalize homelessness can now do it in the open or not at all.
So the fair statement is that the nineteenth century identified the right problem and built an instrument too blunt to solve it.
The Failure Hid Inside the Word “Subject”
A subject is not a measurable unit.
Vehicle registration might be a subject. So might transportation. So might public safety. So might state administration. So might government.
Once a category grows broad enough, nearly any combination of provisions can be described as parts of one legislative purpose. Pull the point of view back far enough and provisions that looked unrelated resolve into a common heading. Pennsylvania’s courts have acknowledged this repeatedly, while also warning that enforcing the rule too strictly would turn judges into supervisors of legislative drafting style.
That leaves the judiciary hunting for a boundary the constitutional text never supplied.
Was this collection of provisions all about gaming? Criminal justice? Public welfare? The operation of government?
A court inclined to uphold a law describes its subject broadly. A court inclined to strike it describes the components narrowly. Both courts can claim, with straight faces, to be applying the same rule.
The doctrine that grew up around this centers on germaneness — whether provisions bear a sufficient relationship to a common purpose. Courts and scholars have found that test vague, malleable, expensive to litigate, and impossible to apply consistently. The leading academic treatment describes the indeterminacy of “subject” as inevitable rather than fixable.
Calling this a failure of judicial courage misses what happened. It is a failure of measurement. The constitutions told judges to enforce one subject without telling them what a subject was, and no amount of judicial nerve produces a unit of measure that the text never contained.
The Enforcement Problem Nobody Mentions
There is a second failure underneath the first, and it deserves more attention than it gets.
Over the last two decades, state courts have used single-subject rules to invalidate laws on firearms regulation, abortion, tort reform, immigration, local minimum wage ordinances, sex offender policy, criminal penalties, and school vouchers.
Look at that list.
Every item on it is politically charged. That is not because legislatures only bundle controversial provisions — they bundle everything. It is because nobody sues over a bundled bill they liked. Single-subject litigation gets brought by the losing side of a policy fight, using a structural rule as the last available weapon.
Which means that in practice, enforcement of the single-subject rule looks partisan almost every time it happens, regardless of whether any particular decision was correct. A judge striking down a firearms law on structural grounds will be described as a judge who wanted to strike down a firearms law. A judge upholding one will be described the same way from the other direction.
The rule therefore burns its own legitimacy every time it is used, and courts know it, which is one honest explanation for why they use it so sparingly.
A line-addressed system takes judges out of that business entirely.
The structure gets enforced at drafting, before enactment, by a rule about form that has no opinion about content. A proposition either states one actor, one operative verb, one condition set, and one legal consequence, or it does not. No court has to decide whether homelessness policy is germane to county accounting, because the two propositions were never permitted to occupy the same vote.
Structural discipline stops being a weapon that only the losers of policy fights can pick up.
The Old Rule Governed Containers
The state constitutions apply their rule to a bill, or to an enacted act.
That is a container.
A bill can satisfy a one-subject requirement and still run hundreds of pages, carry thousands of clauses, delegate dozens of powers, and make a range of policy choices no legislator would support standing alone.
A criminal justice bill may alter sentencing, police authority, prison administration, evidence rules, court funding, and parole. A public health bill may reach hospitals, insurance, licensing, data collection, emergency powers, and school requirements.
One broad heading does not produce one decision. It produces one package.
And the legislator faces the same coercive choice the 1844 drafters were trying to eliminate: accept provisions you oppose to get the one you believe your district needs, or kill the one your district needs to stop the ones you oppose.
That is logrolling. It remains logrolling when every provision can be made germane to a sufficiently expansive topic.
Line voting changes the unit of consent.
Legislatures still deliberate over comprehensive programs. Committees still develop interdependent groups of lines. Negotiation still produces compromise, and no government can operate if every provision is conceived in isolation from every other.
But enactment happens at the operative line.
A legislator votes yea on the deadline, nay on the enforcement mechanism, yea on the appeal right. The record shows all three separately.
The package survives as an organizing device. It stops functioning as a place to hide consent.
The Old Rules Arrived After the Damage
Single-subject restrictions get enforced through litigation, which means after everything has already happened.
The legislature passes the bill. The governor signs it. Agencies begin implementing. Citizens rearrange their affairs around the new requirements. Someone eventually sues.
Years later a court decides whether the parts were sufficiently related, and then either upholds the whole act, invalidates the whole act, or severs the offending provisions while preserving the rest.
Even successful enforcement arrives as demolition.
It does not produce the missing law. It does not identify the correct replacement. It does not force the legislature to reconsider each proposition on its merits. It does not make the next drafting cycle any clearer. It tells a legislature that a package crossed an uncertain constitutional line, long after the moment when knowing where that line sat would have mattered.
A line-addressed system moves diagnosis forward to drafting.
Before a line can be enacted, the drafting system checks whether it has a complete actor, an operative verb, defined conditions, a stated legal effect, valid references, a proper address, and an intelligible relationship to every line it modifies. It also tests whether the proposed line duplicates, contradicts, or silently disables existing law.
None of that decides whether the policy is wise. It decides whether the proposition is structurally complete.
A legislature stays entirely free to enact bad policy. It should not stay free to enact technically indeterminate law without being told exactly where the indeterminacy sits.
The Old Rules Had No Memory
A court strikes a provision. The decision becomes precedent. The legislature may try again.
The legal system itself learns nothing.
No permanent defect record attaches to the failed provision. No structured account survives of which relationship was improper. No mandatory repair goes back to the legislature. A later legislature can recreate the identical problem in different words, and frequently does.
A line-addressed system has memory built into the address.
Every line carries its enactment date, its sponsors, its vote, its prior versions, its dependencies, its judicial constructions, its constitutional challenges, and every certified defect found in its operation.
When a court finds a gap or a conflict, the discovery does not disappear into an opinion. It gets filed against the affected address, classified under the framework developed in the companion brief on judicial structure — bad legal workmanship, a legislative values choice, or a genuinely new question — and placed before the legislature for repair.
The diagnosis follows the line.
Failure becomes maintenance instead of demolition.
A Line Is Not a Topic
Under this design a line is not whatever a legislature places between two paragraph numbers.
It is a structured legal proposition. One actor. One required, permitted, or prohibited action. One stated set of conditions. One legal consequence. One address.
A licensing board shall issue a written decision within thirty days after the evidentiary record closes.
That line performs one job.
A second line defines what closes the record. A third authorizes an extension under stated conditions. A fourth specifies the remedy when the deadline is missed.
Those four lines are related. They belong to the same chapter and serve the same policy. They remain four separately visible propositions.
The old rule asks whether all four belong broadly together — a question with no determinate answer.
The line-addressed system asks questions that do have answers. What exactly does this line command? Who is bound? What facts activate it? What other lines does it depend on? Does it conflict with an existing command? Did each legislator vote for this proposition specifically? What changed when it was amended, and who changed it?
That is the difference between classifying legislation and engineering law.
Riders and Systems Are Different Things, and the Difference Is Measurable
Some policies genuinely require coordinated provisions.
A licensing system needs eligibility standards, an application process, fees, inspection authority, appeal rights, enforcement powers, and transitional rules. Those provisions are not corrupt because they travel together.
This is a real reason courts have resisted strict enforcement. A rule rigid enough to forbid every combination would make coherent legislation impossible. Serious scholarship goes further and questions whether logrolls and riders are as pernicious as reformers assume — some bargaining and packaging is how representative bodies assemble majorities at all, and a system that forbids trading forbids governing.
That objection is correct on its own terms, and the design accepts it.
Line voting does not prohibit structure or coalition. It makes both visible.
A licensing program may contain fifty addressed lines. The system shows that lines one through twelve cannot operate without thirteen through twenty. Legislators see the dependency. The public sees it. A line may declare a condition on its face — effective only upon enactment of a specified line — so that a member who will vote for the cleanup only if the tax passes can make exactly that trade, on the record, at a public address.
Coalition survives. Compromise survives. Concealment does not.
And a rider stops being defined by whether a judge finds it germane to an abstract topic. A rider becomes visible because its address, dependencies, and function show it is not required to operate the structure it was placed inside.
The question stops being whether this feels related. It becomes: what does this line connect to, and what breaks if it is removed?
That question has an answer.
Codification Was Treated as the Exception
Pennsylvania’s constitution forbids bills containing more than one subject and then expressly excepts general appropriation bills and bills codifying or compiling the law. The Commonwealth’s own drafting practice recognizes that substantive changes buried inside codification legislation still raise the underlying problem, because reorganizing existing law and changing policy are different acts.
That exception reveals the assumption underneath the whole old design.
The old system treats codification as something done after lawmaking. Gather the accumulated acts, organize them, reconcile the language, publish the result, and hand it to whoever needs to find something.
This design makes codification the medium in which lawmaking happens.
A legislature does not enact a freestanding pile of text and ask specialists afterward where it belongs. It edits the code directly.
The legislative proposition stops being enact the attached 480-page bill and becomes:
Add line 14.3.7.
Replace line 14.3.9 with the following sentence.
Repeal line 18.6.2.
Change the definition governing lines 18.6.3 through 18.6.11.
Every amendment names the exact object it changes. Every vote produces a new complete version of the code.
No citizen should ever have to set an original statute beside six later acts and perform the amendments in their head to determine what the law says today. The official code displays the law now in force. Its history stays available behind it.
Congress Never Even Tried
One fact belongs in any honest account of this history.
Forty-three states wrote a single-subject rule into their constitutions. The federal Congress has none. It never adopted one, and the amendment that would impose one has been proposed and gone nowhere for years.
So the objection cuts strangely. The critic says the states proved this cannot work, while the institution producing the largest and least readable legislation in the country has never attempted even the weak version.
The states at least identified the disease and reached for a treatment. Washington skipped both steps and has spent a century and a half arguing about the symptoms.
The Thousand-Year Question
No code stays unchanged for a thousand years, and that is not the objective.
The objective is a structure that can be maintained for a thousand years without requiring each generation to excavate the drafting accidents of every generation before it.
A durable code has to assume that language will drift, technology will arrive, institutions will reorganize, moral judgments will move, disputes will emerge that nobody imagined, and lines that look sound today will fail under conditions nobody can predict.
Durability does not come from writing perfect law once. It comes from making law diagnosable and repairable.
A stone wall lasts because individual stones can be inspected, removed, and replaced without demolishing the structure around them. The address stays stable. The current text is visible. Old versions remain preserved. Dependencies can be traced. A court can identify the exact point of failure. A legislature can change that point without hiding unrelated changes around it. The public can see who approved the repair.
That is a modest ambition compared to the alternative, which asks every lawyer, judge, administrator, and citizen to reconstruct a government’s accumulated intent out of centuries of acts, amendments, opinions, regulations, and procedural custom.
What the Failures Actually Teach
The state experiments teach one lesson repeatedly, and it is not that clarity is impossible.
Do not say one subject and leave judges to define subject. Define the legal unit.
Do not regulate the container while leaving the operative provisions bundled inside it. Vote on the provisions.
Do not wait for litigation to discover structural defects. Validate the structure before enactment.
Do not let amendments operate as textual riddles. Change addressed lines directly.
Do not let a court merely destroy defective legislation. Send the diagnosis back through a mandatory repair loop.
Do not treat codification as clerical work performed after politics. Make the code the place where politics has to leave a permanent, traceable result.
Do not promise that every law will be correct. Build a system that can show precisely where it is wrong.
Forty-three states tried to restrain legislative abuse with a sentence. Their failures prove that clarity cannot be commanded vaguely. They do not prove it cannot be built.
Every line simple. Every line addressed. Every relationship visible. Every change traceable. Every proposition voted. Every defect returned for repair.
That structure guarantees nothing about the justice of a society. No arrangement of words can do that.
It can give a just society a legal system solid enough to live inside, and give the society after ours a code it can repair without first becoming an expert in our confusion.
