Release the Brakes, but Never Seize the Wheel: America's Great Presidents and the Three Layers of Democracy

III. The Man Who Kept Every Procedure Lost the Union

Now look at the other end of the list.

Buchanan sits last or second to last in almost every ranking. In December 1860, as the Southern states prepared to secede, his annual message to Congress took this position: secession is unconstitutional, but the federal government also has no power to coerce a state to remain; the Union rests on public opinion, and Congress should decide what to do next.

Procedurally, it is hard to fault a word of it. He did not bypass Congress, did not defy the courts, did not imprison anyone.

By the time Lincoln took office three months later, seven states were gone.

Andrew Johnson, Pierce, and the other perennial bottom-dwellers draw much the same verdict: faced with crisis, they refused to use the power in their hands.

So what the historians' rankings reward has never been "keeping procedure." What they reward is this: when the whole system's survival was at stake, did someone act; and after they acted, did the system survive and come out larger than before.

And that brings the question back to the first layer.

---

IV. The Dividing Line Is Not the Breach, but What Comes After

Return to the three layers, and a decisive fact surfaces: the highest-ranked presidents breached the second and third layers most, yet almost without exception they held the first layer, and often enlarged it.

Lincoln's war was still undecided in the autumn of 1864. On August 23 he wrote a memorandum, folded it, and had every member of his cabinet sign it without reading it. It said: if the election is lost, he would cooperate with the president-elect to save the Union before inauguration day, since the new president would not be able to do so afterward.

In November he won. Two days later he told a crowd that had come to congratulate him: if the rebellion could force us to forgo or postpone a national election, it might fairly claim to have already conquered and ruined us.

Roosevelt likewise held elections on schedule in 1942 and 1944. Jefferson's purchase was ratified by the Senate and funded by Congress. Lincoln's suspension of the writ was ratified by Congress that August and authorized by statute in 1863.

Taken together, the line between a great president and an autocrat is not whether they crossed the line, but three things.

First, breach in the open, not in secret. Lincoln wrote his reasons for suspending the writ into his message to Congress; Jefferson wrote his constitutional doubts into letters to a senator. Nixon also expanded presidential power dramatically, and ranks near the bottom; the difference is that his breaches were designed to hide things from his own voters.

Second, submit the breach to a later verdict, and accept the verdict. Roosevelt's court plan was killed by a Senate his own party controlled, and he did not try again. Jefferson's own theory was exactly this: the president is like a guardian who makes an unauthorized but advantageous investment for his ward, then submits it for ratification afterward, rather than claiming in advance that he had the right.

Third, never touch the first layer. You may release the brakes, but you may not take the wheel out of the passengers' hands.

Measured by these three standards, Jackson turns out to be the odd one on the list. He enlarged the first layer, permanently sacrificed the third, and never submitted his treatment of the Cherokee to any verdict at all. His steady slide down the rankings over the past two decades shows that the historians' own standard is being corrected.

---

V. An Objection That Must Be Answered

At this point a sharp reader will ask: isn't this just victor's justice? The breaches that succeed are called greatness; the ones that fail are called tyranny. Isn't your standard of "open, ratified, elections intact" simply reverse-engineered from the winners?

The objection deserves a serious answer, in three parts.

First, the standard was not derived from winners; it was tested on losers. Roosevelt's court plan was a defeat at the very peak of his power, and he accepted it. That acceptance is itself part of the standard, independent of whether he ultimately won.

Second, the same person can pass and fail at once. Roosevelt's New Deal passes this test; Executive Order 9066 does not. It invoked military necessity while suppressing contrary evidence inside the Justice Department, it was not corrected in time, and the Supreme Court did not plainly say it had been wrong until 2018, in a different case. A standard that can separate the good from the bad within a single president is not one tailored to winners.

Third, Jackson is a counterexample to the standard, not a confirmation of it. If the standard really were victor's justice, then Jackson, who expanded the franchise, won re-election, and built the modern party system, should sit securely near the top. In fact he is falling.

---

VI. The Nineteenth Century Could Draw the Line; the Twentieth Finds It Harder

By now the opening question can be given a fairly complete answer.

The most constructive presidents did indeed do the most damage to the layers of procedure and rights. But a mature constitutional system was never sustained by the absence of such presidents. It is sustained by its ability, after a breach, to do one of two things: ratify it, or draw a line.

In nineteenth-century America this cycle ran fairly completely. Jefferson breached; Congress ratified. Lincoln breached; Congress ratified, and after the war the Supreme Court drew a line in Ex parte Milligan: as long as the civil courts are open, civilians may not be tried by military tribunal.

The twentieth century began to change. In 1948 the political scientist Clinton Rossiter wrote Constitutional Dictatorship, taking Lincoln and Roosevelt as his models, and argued that a democracy handing extraordinary power to one person in a crisis is not necessarily a failure of democracy; what matters is three things: the conditions of entry are clear, the power has a term, and at the end the power is actually handed back.

Twenty-five years later Arthur Schlesinger wrote The Imperial Presidency and corrected Rossiter: power expanded in a crisis rarely returns to where it started, and every president treats the predecessor's tools as a floor rather than a ceiling.

Nearly forty years after that, Jack Balkin and Sanford Levinson argued in 2010 that the modern American problem is no longer a president seizing power temporarily in a crisis, but a Congress that has spent decades depositing vast emergency powers inside the executive branch in advance, where they sit in ordinary times, available to whoever takes office.

The three of them together describe a single curve: the "draw a line" step after a breach has grown later and softer.

---

Closing

Back to the question we started with.

Did America's most constructive presidents do the most damage to traditional democracy? If "democracy" means procedure and separation of powers, or individual rights in an emergency, the answer is yes, and not by a little. Those stains should not be washed away by great outcomes. Interning 120,000 citizens of Japanese descent does not become acceptable because the war was won.

But if "democracy" means the most basic thing of all, that power comes from elections, can be opposed, and can be peacefully replaced, the answer is the opposite. The presidents who were least obedient to procedure were the ones who guarded this one rule most fiercely at the hardest moments. And the most procedurally obedient president watched it collapse.

That, I think, is the real test of a constitutional system. Not whether a president will ever cross the line, but whether, after one does, the system still has the capacity to say "this one we ratify," or "this far and no further."

Releasing the brakes is sometimes necessary.

But the wheel must stay in the passengers' hands.

— Di Yao