FDR Tried to Pack the Supreme Court in 1937 and His Own Party Said No

Imagine you have just won the biggest reelection victory of the century. Forty-six of forty-eight states. Your party holds enormous majorities in the House and the Senate. The voters have told you, in the loudest voice a democracy has, to keep going. And there is one little problem standing between you and everything you want: nine people in black robes who keep reading the Constitution and keep finding that it says no.

What do you do? If you are Franklin Roosevelt in the winter of 1937, you do the thing that seems obvious to a man who has never lost a fight he could not wrap in a slogan. You change the number of people in the room. And here is the part I love, folks. Here is the part every person who casually says "we should add a few seats to the Court" needs to hear. It did not fail because Republicans stopped it. Republicans were a small minority in Washington that year. It failed because Roosevelt's own party looked at the idea, looked at the Constitution, looked at the precedent it would set, and said no.

Let me tell you how that happened, because the story is better than the textbook version and the lesson is bigger.

The landslide and the wall

Start in November 1936. Roosevelt carried every state except Maine and Vermont. He had spent four years building the New Deal, and the voters had just ratified it. But the Supreme Court had been taking the New Deal apart one piece at a time.

In 1935, in Schechter Poultry Corp. v. United States, the Court struck down the National Industrial Recovery Act, and it did so unanimously. Even the Court's liberals could not find the power in the Constitution for Washington to write fair-competition codes for a kosher poultry business in Brooklyn. In January 1936, in United States v. Butler, the Court struck down the Agricultural Adjustment Act's processing tax. In Carter v. Carter Coal Co. that same year, the Court struck down the Bituminous Coal Conservation Act. In June 1936, in Morehead v. Tipaldo, a five-justice majority struck down a New York minimum wage law for women, relying on a 1923 precedent, Adkins v. Children's Hospital.

Four justices were the reliable votes against the New Deal: Willis Van Devanter, James McReynolds, George Sutherland and Pierce Butler. The press called them the Four Horsemen. Three justices, Louis Brandeis, Harlan Fiske Stone and Benjamin Cardozo, usually voted the other way. Chief Justice Charles Evans Hughes and Justice Owen Roberts sat in the middle, and a good deal of American history in those years depended on how those two voted on a given Monday.

Now, I want to be fair to Roosevelt's frustration. A president who runs on a program and wins by that margin is entitled to be annoyed when a handful of unelected judges strike down his laws. That is a real feeling and a legitimate argument. The question was never whether he could complain. The question was what he was willing to break in order to win.

The plan, dressed up as housekeeping

On February 5, 1937, Roosevelt sent Congress what he called a judicial reorganization plan. Read the message and you will find that the word "packing" appears nowhere. Instead, the president talked about the workload of the federal courts and the age of the judges. His proposal was simple on its face. For every federal judge who had served ten years and did not retire or resign within six months after turning seventy, the president could appoint an additional judge. On the Supreme Court, the additions were capped at six, and the Court could never grow larger than fifteen.

Do the arithmetic, because the arithmetic is the whole game. In February 1937, six of the nine justices were already past seventy. Nine sitting justices plus six new appointments equals fifteen. Six new seats filled by one president, on a Court where he had not yet had a single vacancy to fill, would have handed him a working majority of nine out of fifteen without a single justice leaving. Call it what you like, but a plan that gives a president the power to create a majority is a plan to create a majority.

Notice the sales pitch. It was age. It was an overworked bench. It was efficiency. It was the sort of thing a man says when he knows the honest sentence, "I want the Court to rule my way," will not sell in Little Rock or Boise. That should sound familiar to anybody who has watched modern debates about the Court. The respectable-sounding label rarely matches the actual goal.

The fireside chat and the letter that sank it

Roosevelt took his case to the country in a fireside chat on March 9, 1937. He told the audience that the Court had been acting not as a judicial body but as a policy-making body. It is a fair criticism of some decisions, and I have made versions of it myself. But the remedy he proposed was not to appoint better judges as vacancies came. It was to add judges until he had the votes.

The opposition organized in a way the White House did not expect. The leader of the Senate fight was not a Republican. He was Burton Wheeler, a Democrat from Montana, a progressive who had backed Roosevelt in 1932 and had no sympathy for the Four Horsemen. Wheeler argued that the plan would make the Court a creature of whoever happened to hold the presidency, and that a Court packed by a friendly president could be packed by an unfriendly one later. He was right about that, and everybody in the Senate knew it.

Then came the letter. On March 21, 1937, Chief Justice Hughes sent a letter to Wheeler, which Wheeler read at the Senate Judiciary Committee. Hughes wrote that the Court was fully abreast of its work, and that adding justices would not speed the docket but slow it, because more justices means more voices in every conference and more opinions to argue over. The letter was prepared with the knowledge of Van Devanter and Brandeis, and Brandeis, the intellectual hero of the progressive movement, standing behind it blew a hole in the claim that this was about efficiency.

The switch in time

Here is where the story gets interesting, and where historians still argue. On March 29, 1937, the Court decided West Coast Hotel Co. v. Parrish. The case involved Elsie Parrish, a chambermaid in Washington State who sued her employer for back pay under the state's minimum wage law. By a five-to-four vote, with Roberts joining Hughes, Brandeis, Stone and Cardozo, the Court upheld the minimum wage law and overruled Adkins, the case that had carried the day in Morehead less than a year earlier.

On April 12, 1937, the Court followed with NLRB v. Jones & Laughlin Steel Corp., again by five to four, upholding the Wagner Act and the National Labor Relations Board. The Court that had killed the NIRA now accepted that Congress could regulate labor relations in industries affecting interstate commerce. On May 18, 1937, Willis Van Devanter, one of the Four Horsemen, announced his retirement. That gave Roosevelt his first vacancy.

The wits in Washington had a line for the sudden change: "a switch in time that saved nine." The line is usually credited to a humorist of the day, and you will see it attributed to more than one person, so treat the authorship as unsettled. What is better settled is the timing. The justices' vote in Parrish is generally reported to have been taken at their conference in December 1936, well before Roosevelt announced his plan, and the decision was simply held until March. So the popular version, in which Roberts changed his vote in a panic after the plan was announced, does not fit the record well. Roberts had a stated reason for his earlier vote in Morehead: the New York lawyers had not asked the Court to reconsider Adkins, and he said he would not do it on his own.

I think the honest summary is this. Pressure may well have shaped the climate in Washington, and nobody can read a man's mind. But the Court had already begun to turn before the plan was public, and the plan came after the turn, not before it. For conservatives that is a more useful lesson than the legend. It means the threat to the institution was not needed to get a change of course, and it did not even work in the way its authors claimed.

The Senate says no

The Senate Judiciary Committee, controlled by Roosevelt's own party, took up the bill in the spring. In May 1937 the committee reported it adversely, and its report was fierce. It called the plan a measure so dangerous that it ought to be rejected in a way that would make sure nobody tried it again. I will not quote the exact words here, because the report is a public document you can pull up and read for yourself, and it is worth reading.

The bill still had a path on the floor, led by the Senate Majority Leader, Joseph T. Robinson of Arkansas. Robinson was a loyal Democrat, a Roosevelt ally, and widely understood to be in line for a seat on the Court if the plan passed. He worked the floor through the hot Washington summer. On July 14, 1937, he was found dead in his apartment, of a heart attack. The Senate lost its floor manager, and the plan lost its engine.

On July 22, 1937, the Senate voted to send the bill back to the Judiciary Committee, by a lopsided margin, with instructions to drop the Supreme Court provisions. The bill that later came out of Congress, the Judicial Procedure Reform Act of 1937, dealt with lower-court procedures and left the number of Supreme Court seats exactly where it was. Roosevelt signed it in August. Nine justices stayed nine.

And a final irony. Roosevelt lost the fight and still got his Court. He won four terms, and over the years he named eight justices to the bench: Hugo Black, Stanley Reed, Felix Frankfurter, William O. Douglas, Frank Murphy, James Byrnes, Robert Jackson and Wiley Rutledge, and he elevated Stone to Chief Justice in 1941. Time did what the bill could not. A president who stays in office long enough, and who wins elections, fills the Court the old way, by waiting. That is exactly how the Constitution is built to work.

How big has the Court been?

Here is a fact worth carrying to your next family argument. The Constitution does not say how many justices there are. It says there shall be a Supreme Court and gives Congress the power to organize the courts below it. The number has been set by statute, and it has moved.

• 1789: The first Judiciary Act set the Court at six, a Chief Justice and five associates.
• 1801 and 1802: An act passed near the end of John Adams's term would have shrunk the Court by one when the next vacancy came. The new Jeffersonian Congress repealed it in 1802 before it ever took effect.
• 1807: Congress added a seventh seat when it created a new circuit.
• 1837: The number went to nine.
• 1863: It went to ten, with a seat added during the Civil War.
• 1866: Congress passed a law shrinking the Court toward seven by attrition, which was widely understood as a way to keep President Andrew Johnson from filling vacancies.
• 1869: The Judiciary Act of 1869 set the number at nine, where it has stood ever since.

So the count has fluctuated, and every fluctuation was driven by something, whether it was geography, a war, or a fight with a president. Nobody who wants to enlarge the Court needs to be told that Congress has the power to do so. Of course it does. Look at 1863, and look at 1869. The power is not the question.

The strongest case for enlarging the Court, answered

Here is the best version of the argument for adding seats, stated the way its defenders would state it. One, the Constitution gives Congress the power to set the size of the Court. Two, Congress has changed the size several times in American history. Three, a Court that stands in the way of a clear and repeated electoral mandate is, to that extent, undemocratic. Four, therefore adding seats is a legitimate use of an ordinary power.

I will give the first two points to the other side outright. Those are facts. Point three is a value claim, and it is where the argument goes off the rails. Here is why.

First, the Court's job is not to follow the election returns. It is to say what the law is. A Court that votes with the last election has stopped being a court and become a third legislative chamber, and we already have two of those. Second, the changes to the Court's size in the nineteenth century were driven by the growth of the country's circuits or by a fight over a particular president's appointments. The 1866 reduction, aimed at Andrew Johnson, did not last, and Congress put the number back at nine three years later. When Congress has played games with the size of the Court for partisan purposes, the games have not lasted.

Third, and this is the one that should settle it for anybody who thinks past the next election, the weapon does not stay in the hands of the person who first picks it up. If a president can add six justices because he dislikes the Court's decisions, the next president can add six more, and the Court becomes a prize to be fought over in every election instead of a body that the losing side can still trust. That is the argument Burton Wheeler made, and it is the argument that carried the Senate in 1937. A rule is only worth having if you are willing to live under it when the other side holds the pen.

And fourth, look at who stopped it. Roosevelt had just won forty-six states. If there was ever a moment when a power grab over the Court could have passed, this was it. It did not, and it was his own side that said no.

What it means for the argument today

When you hear someone say the Court is out of step and needs to be "reformed" by adding seats, ask three questions. What would be the number, and who decides it? What stops the next majority from doing the same thing? And what will you say when the other side does it to you? The answer to the last one is always the same: nothing. That is why Roosevelt's own senators stopped him.

The other thing to take from this story is that the Constitution already contains a way for voters to change the Court. It runs through the Senate confirmation process, which we cover step by step in Every Step in Confirming a Federal Judge, from Vacancy to Oath. It runs through the elections that decide who holds the presidency and the Senate. It is slower than a statute, and that is the point. If you want the history of how the confirmation fights got to where they are, read From 98 to 0 to Party Lines, a Timeline of Supreme Court Confirmations. If you want to know where the Court got the power to strike laws down in the first place, that is Marbury v. Madison and the Power the Supreme Court Gave Itself in 1803. And if you are curious about how the decisions of the 1930s were reasoned, the commerce clause fight is laid out in A Farmer's Wheat Crop and the Long Fight Over the Commerce Clause.

For readers who want the political side, the party's long exile after 1932 is told in The New Deal Coalition and the Twenty Years the GOP Spent in the Wilderness. And for those of you who think a big majority makes the Constitution optional, see Why the Filibuster Is the Best Friend a Minority Party Ever Had.

The bottom line

• A president who won forty-six states could not pack the Court, because senators of his own party refused. Institutions matter more than majorities, and that is a gift to whichever party is out of power.
• The size of the Court is set by statute and has changed before, so the question is never "can Congress" but "should it." On that, the answer from 1937 and from 1869 is no.
• Remember the test: would you accept the rule if the other side held the pen? If not, you do not believe in a rule. You believe in a result.
• Ask any candidate who talks about adding seats for a number, a limit, and an answer to what happens when the other party does the same.
• Read the Court's own opinions and the Senate Judiciary Committee's 1937 report yourself, and judge the claims against the record.