Madison's Plan for a Government That Fights Itself on Purpose

Washington "can't get anything done," says the fellow at the barbecue, as if that were a bug. It is the specification. Suppose you were hired to design a machine that would be very hard to use for evil. You would not put one lever, in one hand, in charge of everything. You would want the guard, the fence, the kill switch, and the second person standing by the plug, as anybody who has ever owned a table saw can tell you.

James Madison wanted the same thing for the government of the United States, and in the winter of 1788 he explained why in a newspaper essay that every American ought to be able to summarize at the dinner table. It is Federalist No. 51, and it contains the most important sentence in American political science: "Ambition must be made to counteract ambition." Let me walk you through what he meant, where it lives in the text of the Constitution, and why the people who call it "gridlock" when they are winning call it "the system working" when they are not.

Who wrote it and what it was for

The Federalist Papers were 85 essays published under the shared pen name Publius by Alexander Hamilton, James Madison and John Jay, during the fight over whether New York, and the country, would ratify the Constitution that the convention in Philadelphia had signed in September 1787. They were newspaper arguments, written fast, for voters who were nervous about handing power to a new national government. Federalist No. 51 is generally attributed to Madison and appeared in early February 1788.

Keep that audience in mind. Madison was not writing for professors. He was talking to farmers, merchants and lawyers who had just fought a war against a king and were not about to sign up for another one. The question on their minds was simple: what stops this new government from becoming the thing we just beat? Federalist No. 51 is his answer.

Men are not angels

Madison starts with a sentence that is half a joke and half a philosophy: "If men were angels, no government would be necessary." He continues that if angels were to govern men, neither outside nor inside controls on government would be necessary. Then comes the line that every good constitution has to face: in framing a government to be run by men over men, the great difficulty is that you must first enable the government to control the governed, and then oblige it to control itself.

Read that again, because it is the whole job in one breath. A government too weak to control the governed gives you chaos, which the founders had just tasted under the Articles of Confederation. A government that cannot be made to control itself gives you tyranny, which they had tasted under George III. The art is doing both.

Madison does not pretend the voters alone will solve it. He concedes that "a dependence on the people is, no doubt, the primary control on the government," and then adds that experience has taught mankind the necessity of "auxiliary precautions." Elections are the main check. But elections come every two or four or six years, and a lot can happen in between. So you build in more.

The design: ambition against ambition

Here is the engine of the argument. If you cannot rely on angels, rely on human nature itself. Give each part of the government the means and the motive to resist the others. Madison says the great security against a gradual concentration of the several powers in the same department is giving to those who administer each department the necessary constitutional means and personal motives to resist encroachments of the others. And then he says that the interest of the man must be connected with the constitutional rights of the place.

That is a very conservative insight about people. It does not ask a senator to be selfless. It assumes that a senator wants to protect the Senate's prerogatives because those prerogatives are his own. It assumes a president wants to guard the presidency. The Constitution does not try to remove ambition from politics. It aims it. Think of it as judo: use the other fellow's energy against him.

Where it lives in the text

Madison was describing a design, but you do not have to take his word for the design. Pull out the Constitution and count the checks yourself. Here are six that any citizen can verify in a few minutes.

1. Two houses. Article I, Section 1 vests all legislative power in a Congress of two chambers. A bill must pass both. The House has two-year terms and the Senate has six, so the two bodies answer to different clocks, and Article I, Section 7 requires revenue bills to start in the House.
2. The veto and the override. Under Article I, Section 7, a bill that passes Congress goes to the President. He can sign it or veto it, and Congress can override the veto only by a two-thirds vote in each house. The executive checks the legislature, and the legislature, with a supermajority, checks back.
3. The power of the purse. Article I, Section 9 says no money shall be drawn from the Treasury but in consequence of appropriations made by law. The President can propose, but only Congress can fund.
4. Appointments and treaties. Article II, Section 2 lets the President nominate judges, ambassadors and other officers, but only with the advice and consent of the Senate, and a treaty needs two-thirds of the senators present.
5. Impeachment. The House has the sole power of impeachment, the Senate has the sole power to try it, and conviction requires two-thirds of the members present. One chamber accuses and the other judges.
6. Judges with tenure. Article III says federal judges hold their offices "during good Behaviour" and their pay cannot be reduced while in office. A judge who does not need the next election can rule against the popular side.

That is only the horizontal check, branch against branch. There is a vertical one too, and Madison was explicit about it. In America, he says, the power surrendered by the people is first divided between two distinct governments, the national and the state, and then the portion allotted to each is subdivided among distinct and separate departments. Hence what he calls a "double security": the different governments will control each other, at the same time that each will be controlled by itself. The Tenth Amendment, which reserves to the states or the people the powers not delegated to the national government, is the written form of that idea. For the practical consequences of the state side, see Three Towns, Three Rule Books and the Case for Letting States Differ.

The part people skip: factions

Federalist No. 51 has a second argument, and it is the one that links to its more famous neighbor, Federalist No. 10, published in November 1787. Madison defines a faction in No. 10 as a number of citizens, whether a majority or a minority of the whole, united by some common passion or interest adverse to the rights of other citizens or to the permanent and aggregate interests of the community. He is blunt that you cannot abolish factions without abolishing liberty, which is like abolishing air because fire uses it up. So the problem is to control their effects.

His proposed cure in No. 10 is size. A small republic can be captured by one dominant faction. A large one, he argues, contains so many interests and sects that it is hard for any single majority to form and harder still to act in concert against the rest. Federalist No. 51 carries the same idea back into the structure: in a society with many interests and many sects, the rights of individuals and minorities are secure because a combined majority becomes less likely. He puts the stakes bluntly. If a majority were united by a common interest, the rights of the minority would be insecure. And he says justice is the end of government, and the end of civil society.

Notice what that does to the usual talking point that "the majority should rule, full stop." Madison is not against majorities. He is against majorities that are in a hurry and unconstrained. The Constitution's answer is to make majorities earn their power: win the House, then the Senate, then the signature, then survive the courts, and do it again at the next election.

The strongest case against Madison

I always want to state the other side at its best, so here it is. Critics say the design produces paralysis. A president promises change, wins a landslide, and then spends two years being told no by a Senate committee. Voters get angry because nothing they voted for seems to happen. Worse, they say, power does not vanish when it is divided. It just migrates to the places that are hardest to see: agencies, courts, unelected staff. A system built to stop Congress and the President from acting together can end up handing the real decisions to a bureaucracy that answers to no ballot.

That objection has teeth, and conservatives of all people should admit it. The growth of the administrative state is partly a story about what happens when the branches decide that the hard work of passing laws is too hard. Congress writes a vague statute, the executive fills in the details, the courts defer, and nobody who faces the voters ever has to own the result. I would argue that is not a flaw in Madison's design. It is what happens when the branches stop playing the roles he assigned them. The remedy is for Congress to behave like Congress, a point we make at more length in Congress Should Write the Laws, and Other Radical Ideas From Article I.

And on the paralysis charge, I have two answers. First, slowness is a feature of a system that is supposed to protect you from a bad idea as much as it delays a good one. A law that cannot pass without persuading more than one body and more than one party tends to be a law that more people can live with. Second, gridlock is rarely total. Congress passed major legislation in divided government many times; it just required the parties to talk to each other. The design does not forbid action. It demands agreement.

Both parties love it, when they lose

Now the fun part. The surest test of whether somebody believes in the separation of powers is whether he believes in it when it is inconvenient. Almost nobody does, and that is the point Madison was making: nobody should be trusted with unchecked power, including you and including me.

Consider the filibuster. In November 2013, the Senate's Democratic majority, led by Harry Reid, changed the Senate's precedents so that most presidential nominees other than Supreme Court justices could be confirmed by a simple majority. Republicans howled. In April 2017, the Republican majority extended the same change to Supreme Court nominees, in order to confirm Neil Gorsuch, and Democrats howled. Each side called the other's move a betrayal of the Senate's traditions, and each side did exactly the same thing when it had the votes. For the longer history of that rule and what it protects, read Why the Filibuster Is the Best Friend a Minority Party Ever Had.

Consider the states. When one party holds the White House, attorneys general from the other party tend to turn into constitutional scholars overnight, filing suits about the limits of executive power and the sovereignty of the states. When the White House changes hands, the same offices swap briefs and the same arguments, with the names reversed. Consider the courts: the party that is losing in Congress discovers a deep respect for judicial review, and the party that is losing in court discovers a sudden interest in term limits. It is a human comedy, and Madison would have recognized every actor.

The honest conclusion is not that everybody is a hypocrite, though plenty are. It is that the structure works precisely because it does not require anybody to be principled. The minority always has a reason to defend the walls. That is ambition counteracting ambition, in real time, on the evening news.

What a family should take from this

Three practical lessons, for people who have a job and a mortgage and no desire to read the Federalist Papers in the original.

1. Judge a proposal by the power it creates, not just the result it promises. Before you cheer a new authority for the President or an agency, picture the person you trust least holding it. If the answer makes you queasy, you have found a flaw.
2. Treat the slow parts as friends. The Senate's six-year terms, the veto, the courts and the states are not obstacles to your agenda. They are the reason your agenda survives the next election.
3. Hold your own side to the rule. The best test of principle is whether you apply the same standard to the person you voted for. If an executive overreach is wrong when the other party does it, it is wrong when yours does it.

If you want to see how the other founding documents and arguments line up with Madison's, The Founding Ideas in Order, From Magna Carta to the Bill of Rights lays out the sequence, and Madison Built a Government Where Ambition Fights Ambition, and It Works looks at the same design through the lens of the courts. For a tour of every article of the text, try A Walkthrough of the Constitution You Can Finish in One Evening. If you want the broader picture of how the party has handled these limits, our /politics/republicans/ page tracks the party's record.

The bottom line

• Madison's premise was that men are not angels, so the Constitution does not ask officials to be good. It makes them rivals, and rivals watch each other.
• You can check every claim above yourself: Article I for the two houses, the veto and the purse, Article II for appointments and treaties, Article III for tenure, and the Tenth Amendment for the states.
• The same rules that frustrate your side when it is winning protect your side when it is not. Defend them in both seasons.
• Next time you hear "gridlock," ask who benefits from calling a safeguard a malfunction, and whether they would say the same thing if the other party were in charge.
• Talk to your family about the habit of mind, not the party: before backing any new power in Washington, ask who holds it after the next election.