Quilling illustration for Federalist Nos. 47-51
Federalist Nos. 47-51

The Separation of Powers

Generated summaryMadison explains what actually keeps power divided among the branches: not a wall of paper, and not popular votes to referee disputes, but a structure that gives each branch the means and the motive to defend itself against the others.

Federalist 47

Generated summaryMontesquieu never meant the branches to be wholly disconnected. His point, confirmed by Britain’s own constitution and every state’s, is only that no branch may hold the whole power of another.
Key points
  • The charge that the Constitution violates the separation of powers fails; the maxim has been misunderstood.
  • To accumulate all powers, legislative, executive, and judiciary, in the same hands is the very definition of tyranny.
  • Montesquieu, the authority always cited, drew the maxim from a British constitution whose departments are by no means wholly separate.
  • His meaning: liberty is subverted only where one department holds the whole power of another, not where they share partial agency.
  • Every state constitution proclaims the maxim yet mixes the departments in practice.
  • The convention’s plan violates it in no point, forbidding only one whole department from exercising another’s powers.

Having reviewed the government’s general form and its mass of power, I turn to its particular structure. My subject is how that power is divided among its parts. One of the chief objections comes from respectable opponents. The Constitution, they say, breaks the maxim that the legislative, executive, and judiciary departments must be kept separate. It so blends them that some parts risk being crushed by the weight of others.

No political truth rests on higher authority than this objection’s maxim. All powers, legislative, executive, and judiciary, in the same hands is the very definition of tyranny. That holds whether the hands are one, a few, or many, hereditary, self-appointed, or elective. Were the Constitution chargeable with that, or with a dangerous tendency toward it, no argument could save the system. I am confident the charge cannot be supported. The maxim it relies on has been entirely misunderstood. We must first find the sense in which liberty requires the three departments to be separate.

The authority always cited here is Montesquieu. If he did not discover the precept, he displayed it best. He treated the British Constitution as the standard, or, in his own phrase, the mirror of political liberty, and set its principles down as elementary truths. To fix his meaning, let us go back to that source.

On the slightest view, the British departments are by no means wholly separate. The executive magistrate forms part of the legislative authority; he alone makes treaties, which within limits carry the force of law. He appoints every judge, and consults the judges as a council. One branch of the legislature is a council to the executive. That branch is also the sole judge in impeachments and the highest court of appeal.

From these facts his meaning is clear. There is no liberty, he wrote, where the legislative and executive powers unite in one person or body. Nor is there liberty where judging is not separated from them. He did not mean that the departments may have no partial agency in, or control over, one another’s acts. He meant that where the same hands hold the whole power of two departments, a free constitution is subverted.

That is not among the vices of the constitution he examined. The king cannot make a law, only veto one; he cannot judge, only appoint the judges. The judges exercise no executive or legislative power. The legislature does no judicial act, though it may remove judges and one branch is the court of last resort. It does no executive act, though one branch is the supreme executive magistracy.

His reasons confirm it. Where the legislative and executive unite, he says, there can be no liberty. One may fear that the same monarch or senate will enact tyrannical laws to execute them tyrannically. Join judging to the legislative, and the judge becomes the legislator; life and liberty lie open to arbitrary control. Join it to the executive, and the judge might act with all the violence of an oppressor.

Look into the state constitutions. Despite the emphatic terms of this axiom, not one keeps the departments wholly separate. New Hampshire, whose constitution came last, seems to have seen that avoiding all mixture is impossible and unwise. It kept the powers only as separate as a free government admits. Its Senate tries impeachments, and its President heads the executive yet presides in the Senate with a casting vote.

Massachusetts states the article less pointedly. No department, it declares, shall ever exercise another’s powers. That matches Montesquieu as explained, and the convention’s plan nowhere violates it. The plan forbids only a whole department from taking another’s powers. Yet that constitution itself admits a partial mixture. Its Senate tries impeachments of executive and judicial officers, and its legislature appoints several officers each year. Appointment is in its nature executive, so on this point the framers broke their own rule.

I pass over Rhode Island and Connecticut. They were framed before the Revolution, before this principle drew notice. New York makes no declaration. Yet it was plainly framed with an eye to the danger of blending the departments. It still gives the executive, and even the judiciary, a partial control over the legislature. Its council of appointment joins legislators with the executive in naming officers, executive and judicial.

New Jersey has blended the powers more than any state so far. Its governor is chosen by the legislature. He is chancellor, sits on the Supreme Court of Appeals, and presides with a casting vote over one legislative branch. In Pennsylvania the legislature dominates the yearly vote for president. With an executive council he appoints the judiciary. He also forms a court of impeachment for all officers, judicial as well as executive.

In Delaware the chief magistrate is elected each year by the legislature. With six others the legislature names, he forms the Supreme Court of Appeals and helps appoint the other judges. The legislature appoints the chief executive officers, and one of its branches forms a court of impeachment. All officers may be removed on its address.

Maryland adopts the maxim flatly: the powers ought to be forever separate and distinct. Yet its executive is chosen by the legislature, and its judges by the executive. Virginia is more pointed still: no person shall exercise the powers of more than one department, except that county justices may sit in either House. Yet its legislature appoints the chief magistrate and his council. It also fills the principal executive and judicial offices.

North Carolina declares these powers forever separate. Yet its legislature appoints the executive chief, and the principal officers of both that department and the judiciary. In South Carolina the legislature elects the executive. It appoints the judiciary too, down to justices of the peace and sheriffs. Georgia declares the departments separate. Yet its legislature fills the executive department, holds the final pardon, and names even justices of the peace.

In citing them I do not defend the state plans. They bear strong marks of the haste, and stronger of the inexperience, in which they were framed. In some the principle was broken by too great a mixture of powers. In none was real provision made to keep in practice the separation drawn on paper. My point is only this: the charge against the Constitution is warranted neither by the maxim’s author nor by the sense it has held in America.

Federalist 48

Generated summaryWords on parchment cannot hold power in place. The legislature draws all power toward itself, and Virginia’s and Pennsylvania’s experience shows the other departments need real defenses, not a line on paper.
Key points
  • Marking the departments’ limits on parchment is not enough; each needs a constitutional control over the others.
  • The legislative department everywhere extends its sphere, drawing all power into its impetuous vortex.
  • Our founders watched only the hereditary executive and forgot the danger of legislative usurpation.
  • In a representative republic it is the legislature’s ambition, not the executive’s, that the people must fear.
  • Its extensive, ill-defined powers let it mask encroachments and control the pay of the other departments.
  • Virginia under Jefferson and Pennsylvania’s Council of Censors both show the legislature overrunning its bounds.

The last paper showed that this maxim does not require the legislative, executive, and judiciary departments to stand wholly apart. Each must have some constitutional control over the others. Without that, the separation the maxim asks for can never hold in practice. All agree that no department should wield another’s powers, or hold an overruling influence over them. Power by nature reaches past its bounds, and must be held inside them. The hard problem is a real guard for each against the others.

Is it enough to mark the bounds in the constitution and trust these parchment barriers? Most American constitutions relied chiefly on them, but experience shows their strength was greatly overrated. The weaker departments need a firmer defense, for the legislature everywhere draws all power into its impetuous vortex. Our founders seem to have watched one danger only: the all-grasping prerogative of a hereditary magistrate. They seem never to have recalled the danger from legislative usurpations. These gather all power into the same hands and lead to the same tyranny.

Where a hereditary monarch holds wide prerogatives, the executive is justly the source of danger. But in a representative republic the executive is closely limited in reach and term. The legislature draws bold confidence from its supposed influence over the people. It is large enough to feel a crowd’s passions, yet not too large to pursue them by means reason prescribes. Against the ambition of this department the people should exhaust every precaution.

The legislature’s powers are wider and harder to bound, so it can mask its encroachments under complex, roundabout measures. The executive is narrower and simpler, and the judiciary is marked by clearer landmarks. A usurpation by either would show itself at once and fail. The legislature alone reaches the people’s pockets, and holds full or leading influence over the pay of those who fill the other departments. That dependence makes encroachment easier still.

I appeal to our own experience, and take two states backed by two authorities beyond objection. The first is Virginia, whose constitution declares that the three great departments must not be mixed. My authority is Mr. Jefferson, its own chief magistrate. He writes that all the powers of government there fall to the legislative body, and that is precisely despotic government. Many hands are no relief: one hundred and seventy-three despots would oppress as surely as one. Nor does our choosing them help. An elective despotism was not what we fought for. We fought for powers so divided and balanced that none can pass its legal limits unchecked.

He adds that Virginia separated the departments, but put no barrier between them. The judges and the executive depend on the legislature for their pay, and some for holding office. So if the legislature takes their powers, no protest is likely, and none that is made can work. It can cast its acts in the form of Assembly law, which binds the other branches. It has often decided rights belonging to the courts, and its command of the executive is becoming a habit.

My other example is Pennsylvania, and my other authority its Council of Censors of 1783 and 1784. That body was charged to ask whether the constitution had been kept whole, and whether either branch had overstepped. From facts both sides on the council mostly granted, it appears the legislature had broken the constitution openly and often. It passed public bills without printing them first, as the constitution requires. It violated trial by jury, assumed powers never delegated, and usurped executive powers. It varied the judges’ pay, which must be fixed, and drew cases belonging to the courts into legislative hands.

Some of this may be charged to the war, but most are the natural shoots of an ill-framed government. The executive was not innocent of frequent breaches either, though three things should be said. First, a great part sprang from the needs of the war, or were urged by Congress or the commander-in-chief. Second, most of the rest followed the known wishes of the legislature. Third, Pennsylvania’s executive has so many members that it looks like a legislative assembly. No member answers alone for the body’s acts, so it ventures beyond its powers more freely than a single hand.

The conclusion I am warranted in drawing from all I have shown is plain. A mere line drawn on parchment, marking the constitutional limits of the several departments, is not a sufficient guard. It cannot stop the encroachments that gather all the powers of government into the same hands. Such a concentration is tyranny.

Federalist 49

Generated summaryJefferson would call a popular convention whenever a branch is thought to overstep. Such appeals would unsettle the government, put passion above reason, and favor the legislature, which could fill the convention with its own men.
Key points
  • Jefferson’s plan of referring constitutional disputes to popular conventions is no fit standing remedy for encroachments.
  • His proposal: when two branches agree by two-thirds, a convention is called to correct breaches.
  • It suits republican theory, for the people alone are the source of power and no department can judge its own bounds.
  • Yet frequent appeals would strip the government of the veneration that time confers, and which it needs for stability.
  • They would rouse the public passions and endanger the public peace.
  • Above all, the legislature would dominate any appeal, pleading and judging its own cause.
  • The passions, not the reason, of the public would decide; but only its reason ought to govern.

The author of the Notes on the State of Virginia added a draft constitution. One precaution in it is perhaps wholly his own, and on it he finally relies to shield the weaker departments from the stronger. Two of the three branches, each by two-thirds of its whole number, may judge a convention needed. It shall then be called to alter the constitution or correct breaches of it.

The people are the only rightful source of power, and every branch holds its power under a charter from them. So republican theory favors going back to them, not only to enlarge, reduce, or remodel the powers of government, but whenever one department encroaches on another. The departments are co-equal under one commission, so none can claim a higher right to fix the boundaries. Only the people, who granted it, can say what it means and enforce it.

The reasoning has force: a constitutional road to the people should stay open for great and extraordinary occasions. But insurmountable objections bar it as a standing rule. First, it does not reach two departments combining against the third. The legislature has many means of working on the motives of the others. It might win over either of them, or even a third of its own members, leaving the last nothing. I do not press this; it strikes at the principle’s form, not the principle.

A second objection lies in the principle itself. Every appeal implies some defect in the government. Frequent appeals would largely strip it of the reverence that time confers. Without that, even the wisest and freest governments might lack stability. Governments rest on opinion, and opinion grows firmer with numbers, firmer still when the examples are old. Even the most rational government gains from the community’s prejudices.

A graver objection is the danger of stirring the public passions and disturbing the peace. Our revisions have worked, but such experiments are too delicate to repeat needlessly. Every existing constitution was framed amid a danger that held down the passions hostile to order. Confidence in patriotic leaders silenced the usual disagreement. Zeal for new forms sprang from anger at the old government, and no party spirit touched the changes. Nothing ahead gives us the same security.

The greatest objection is that such appeals would not keep the constitutional balance. Republican governments tend to enlarge the legislature at the others’ expense, so the appeals would come from the executive and judiciary. Their members are few, known to only a small part of the people. Legislators are numerous and live among the people, and blood, friendship, and acquaintance bind them to the most influential part of society. The other side can hardly have an equal chance. Worse, the legislative party would probably judge its own cause, for the influence that elected its members would win them convention seats. The convention would consist chiefly of men who had been, were, or hoped to be of the department on trial.

Sometimes an appeal might come in circumstances less unfavorable to the executive and judiciary, as when a usurpation is too flagrant to disguise. The decision might then lean less toward the legislative party, but it could never be expected to turn on the true merits. It would be bound up with old parties, or with parties born of the question. And the very men who had pushed or opposed the measures would pronounce it. The passions, not the reason, of the public would sit in judgment. But the public’s reason alone ought to control the government, and its passions ought to be controlled by it.

We found in the last paper that mere declarations cannot keep the departments within their legal rights. Occasional appeals to the people would be neither proper nor effective. How far the plan’s other provisions might answer, I do not examine. Some rest on sound political principles, and all show singular ingenuity and precision.

Federalist 50

Generated summaryReviews on a fixed schedule fare no better. Pennsylvania’s Council of Censors split along old party lines, and proved that a scheduled review turns on faction, not calm judgment.
Key points
  • Periodical appeals to the people at fixed intervals are no better remedy than occasional ones.
  • Short intervals leave measures recent and heated; long ones let abuses root deeply before any review.
  • A distant prospect of censure scarcely restrains power from present excesses.
  • Pennsylvania’s Council of Censors actually tried this, its members the very party leaders whose measures were under review.
  • Split into two violent parties, it proved passion, not reason, governs such bodies, and the remedy fails.

Someone may urge appeals to the people at fixed times, not as occasions arise, to prevent and correct breaches of the Constitution. I judge them only as a way to keep each department within bounds, not to change the Constitution. Fixed times seem nearly as bad as single occasions. Short gaps leave measures fresh and heated. Long gaps leave recent measures the same, and a calmer look at older ones brings offsetting drawbacks. First, a far-off threat of blame barely checks present excess. Second, the abuses often do their harm before the remedy comes. Third, where they do not, the abuses are deep-rooted and hard to remove.

This plan has actually been tried. Pennsylvania’s Council of Censors met in 1783 and 1784. Its charge: to ask whether the constitution had been broken, and whether the legislative and executive departments had encroached on each other. A single trial does not settle the point, yet it bears out my reasoning. First, some of the council’s most active members led the parties already in the state. Second, those same men had shaped both departments in the period reviewed, backers or foes of the very measures now tested. Two had lately been vice-presidents of the state, and one speaker of the assembly.

Third, every page of their record shows the council split into two fixed and violent parties, as they admitted. On every question, however small or unrelated, the same names stand opposed in the two columns. Any fair observer may infer that passion, not reason, ruled. Men who reason coolly differ on some questions; men ruled by one passion agree.

Fourth, the council’s decisions were not surely right; in some cases they may have misread the departments’ limits instead of enforcing them. Fifth, I have never known those decisions, right or wrong, to change the practice built on the legislature’s own reading. So this body proves the disease by its inquiries, and the remedy’s failure by its example. Nor does that state’s party rage excuse it. Will any state be free of parties at a future seven-year mark? We should neither expect nor wish it: the end of parties means either a general alarm for public safety, or the end of liberty.

Nor would it help to bar from these reviews everyone who served in government during the period. The task would probably fall to men of lesser ability, hardly better fitted. Even if they had no hand in the measures examined, they would likely belong to the parties behind them, and be chosen under their banner.

Federalist 51

Generated summaryThe real guard is built into the structure: each department gets the constitutional means, and the personal motive, to resist the others, so ambition checks ambition. Splitting power between the federal and state governments adds a second guard.
Key points
  • The security for separation must come from within, so arranging the government’s parts that each keeps the others in place.
  • Each department needs a will of its own, its members drawn independently and not dependent on the others for salary.
  • Give each the constitutional means and personal motives to resist encroachment: ambition must be made to counteract ambition.
  • If men were angels, no government would be necessary; government must first control the governed, then control itself.
  • Dependence on the people is the primary control, but experience teaches the need for auxiliary precautions.
  • In a republic the legislature predominates, so divide it into branches and strengthen the weaker executive.
  • America’s compound republic splits power between two governments, then each among departments: a double security for the people.
  • The extended republic’s many interests and sects keep any unjust majority from combining against the minority.

What can we finally rely on to keep each department in its place? Outside checks have proved inadequate; the cure must come from the government’s own structure. I will hazard a few observations on how its parts may check each other. All agree that liberty needs the powers kept apart, at least to a degree. So each department needs a will of its own, and little hand in choosing the others. Every office would then come from the people through channels that never meet. Cost and trouble force a few departures, chiefly for judges: they need special skill, and permanent tenure ends dependence.

If the legislature holds their pay, the executive and the judges are independent in name only. Against a slow gathering of power, arm each department with the constitutional means, and the personal motives, to resist the others. Ambition must counteract ambition. If men were angels, no government would be necessary. In a government of men over men, you must first enable it to control the governed. Then you must oblige it to control itself. Dependence on the people is the primary control, but experience teaches the need for further guards.

Rival interests make up for the lack of better motives throughout human affairs. The state’s supreme powers need that prudence no less than the lesser offices. Yet no department can be armed equally, and in a republic the legislature is bound to be the strongest. The remedy is to divide it into branches, kept apart by different modes of election and action. The executive, being weaker, may need strengthening. An absolute veto, its seeming defense, would be neither wholly safe nor enough alone. Perhaps a qualified tie to the weaker legislative branch could supply that defect.

If these principles are just, as I persuade myself they are, the state constitutions bear the test far worse than the federal one. Two further points belong to our federal system. First, a single republic gives all the power the people surrender to one government. Dividing that government into separate departments guards against usurpation. Ours divides the power twice: between two governments, then among departments in each. So the people’s rights have a double security: the governments control each other, and each controls itself.

Second, a republic must guard the people from its rulers and from one another. Classes of citizens differ in interest, and a united majority leaves the minority unsafe. Only two remedies exist. One is a will independent of the majority, and so of society itself: hereditary or self-appointed rule. That security is shaky: such a power may back the majority’s injustice as readily as the minority’s right. The other takes in so many classes of citizens that an unjust majority combination grows very unlikely, if not impracticable. Our federal republic takes this second way. All its authority comes from society. Yet society breaks into so many parts and interests that minorities have little to fear from majority combinations.

Civil rights rest on the same ground as religious rights. That ground is many interests in one case, many sects in the other. Security in both grows with the country’s size and the number of people under one government. So a proper federal system deserves every sincere friend of republican government. Break the Union into smaller confederacies and oppressive majorities grow easier, every class of citizens less safe. The only other security, an independent part of the government, must rise.

Justice is the end of government, and of civil society. Where a society’s forms let the stronger faction unite and oppress the weaker, anarchy rules as in a state of nature. There even the strong, unsure of their safety, accept a government that shields the weak too. So the stronger factions come in time to want a government that protects all parties.

Our extended republic holds many interests, parties, and sects. A majority of the whole could seldom unite except on justice and the general good. The minority runs less danger, so there is less pretext for a will above society. The larger the society, within a practical sphere, the fitter it is to govern itself. And happily for the republican cause, a wise mixture of the federal principle can carry that sphere very far.