The seat outlives the argument

The case for judicial independence is one of the strongest in political design, and nothing here is an argument against it. A judge who can be removed for an unpopular ruling is not a judge; they are a very slow-moving official of whoever can remove them. The people who built life tenure were solving a real problem, and they solved it. That part worked.

What they were not solving is the problem we now have, because it did not exist yet. When the arrangement was written, a lifetime appointment was a fairly short instrument, and the numbers say so plainly: across the Court's history the average length of service has run under fifteen years, while since 1970 it has run to roughly twenty-six. Calabresi and Lindgren, working the seats vacated between 1971 and 2006, put the average at 26.1 years. The record is 36 years and 209 days, held by William O. Douglas; the longest-serving justice currently sitting, Clarence Thomas, passed 34 years and 292 days this August. Nobody designed for that, because nobody could reasonably imagine it. Medicine roughly doubled the length of the term without anyone voting on it.

Independence and unaccountability arrived in the same box. They are not the same thing, and the fact that we have never separated them is an accident of drafting rather than a principle anybody defends out loud. Independence means a ruling cannot be punished. Unaccountability means the reasoning behind it is never required to answer to anything — not to a later court, not to a successor, not to the person who held the view before the holder changed their mind. Run the dial above and the distinction becomes physical: the bars are real tenures in real seats, and the eighteen-year line is crossed so routinely that crossing it stopped being remarkable. Everyone knows that gap exists. No part of the machine registers it.

A vacancy is a coin flip, and the coin decides more than an election does. This is the part that ought to bother anyone who values order, because it is the part that is genuinely arbitrary. Whether a given administration reshapes the bench for a generation depends on actuarial luck: who happens to fall ill, and when, and under whom. The Congressional Research Service counts a vacancy arising on average every two years since 1791 — but only every 3.1 years since 1980, and the gap between the Stevens retirement in 2010 and the Scalia death in 2016 ran 2,054 days, the fifth longest in the Court's history. Before 1900, 59 percent of vacancies opened because a justice died in the chair. That is not a check, a balance, or a considered constraint on power. It is chance, and chance is the one thing an institution is supposed to remove from the exercise of authority. When the stakes of a single death are that high, everything downstream distorts — appointments skew younger to buy decades, confirmation becomes total war because it is the only bite anyone gets, and a justice's retirement becomes a strategic act rather than a personal one.

The strongest counter-argument is that any change is a door, and doors get walked through. Take it seriously, because it is correct. Every mechanism proposed to loosen the grip of a tenured bench is also a mechanism a future majority can use to break it. Term limits can be dressed up as reform and deployed as a purge. Expansion is worse. If the reader's instinct is that the current arrangement is bad but the alternatives are how you lose a court entirely, that instinct is well-founded and it is the reason nothing has happened in two centuries.

But 'reform is risky' is an argument for the *design* of the reform, not against ever making one. The distinction that does the work is prospective versus retrospective. A change that affects only seats filled after it passes takes nothing from anyone sitting, hands nobody a present-tense advantage, and cannot function as a purge because there is no one to purge. It is slow enough to be boring, which in constitutional design is the highest available compliment.

The shape that keeps independence and drops the lottery is a long, single, staggered term. Eighteen years, one seat every two years, no reappointment and no removal — a judge still cannot be punished for a ruling, still owes nothing to whoever appointed them, and still outlasts several administrations. This is not our invention, and it is not a partisan one: Paul Carrington and Roger Cramton put it in draft legislation as the Supreme Court Renewal Act of 2005, Steven Calabresi and James Lindgren argued the same eighteen-year staggered term in 2006, and John Roberts was on record for term limits before he was appointed chief justice. Carrington's point is worth repeating — the Constitution never says 'life tenure' at all; it says judges hold office 'during good Behaviour'. What disappears is the actuarial coin flip: every president gets the same two appointments, no death changes the arithmetic, and the temperature of a confirmation drops because it is a scheduled event rather than a once-a-decade emergency. All three authors would let a justice leaving the seat sit on other federal courts until retirement — the judge leaves the bench, not the law.

This is not an exotic idea imported from somewhere that does not share our assumptions. Almost every constitutional court built in the last century — by drafters who studied ours closely, and admired it — took the independence and declined the life term. Germany's Federal Constitutional Court seats its judges for a single twelve-year term with no re-election and a hard retirement at 68. France's Constitutional Council runs nine-year non-renewable terms with a third of the seats turning over every three years, which is the staggering principle in plain sight. Even the United Kingdom, whose court is the closest cousin to ours, retires its justices at 75. They chose fixed terms or retirement ages precisely because they wanted the protection without the lottery. The inherited arrangement encodes a real reason, the reason is independence, and independence is the part worth keeping.

The honest conclusion is smaller than the subject deserves and truer than the alternatives. The problem is not that the bench is old, and it is certainly not that the people on it are acting in bad faith. The problem is that we have built an institution where the single most consequential variable is how long a human body lasts, and then declined to look at it. President Biden's own commission on the Court delivered a unanimously approved 300-page report on 7 December 2021 and declined to take a position on term limits at all — which is what happens when every proposal arrives wearing a team jersey. Take the jersey off and it is a scheduling problem.