ANALYSISApril 13, 2026
Should the death penalty be ended in the United States?
The United States executed 47 people in 2025 — the highest annual total since 2009 — driven largely by a surge in Florida and the Trump administration's revival of federal executions after Biden had commuted 37 federal death row sentences in late 2024. At the same time, public support for capital punishment fell to a 50-year low of 52%, new death sentences reached historically low levels at just 23, and abolition bills have been introduced in Congress.
When the government executes a person, is it delivering justice — or committing the same irreversible act it's punishing? And if even one innocent person has been killed by the state, does the whole system forfeit its moral legitimacy?
Not familiar with this story? Get context →
Error rate makes system indefensible
C
One exoneration for every 8.3 executions is not a margin of error that any institution operating with irreversible consequences should be permitted to sustain. Conservatives who distrust government competence in healthcare or education cannot coherently exempt the state from that skepticism when the output is a corpse. The irreversibility is not a detail — it is the entire problem.
L
We agree on the number, but notice what the conservative framing does: it treats this as a fixable government competence problem rather than what it is. Cameron Todd Willingham was executed in 2004 for a fire that independent investigators later determined was likely accidental — that is not a bureaucratic inefficiency, it is the state committing the crime it was punishing. Fifty years of 'higher standards' have produced the rate you're quoting.
C
Willingham is the strongest case abolitionists have, and it deserves weight — but it is one documented likely wrongful execution, not evidence that the standard can never be met. The McVeigh execution, guilt established beyond any conceivable doubt, shows the standard is achievable; the question is whether we enforce it instead of treating it as optional.
L
If fifty years of reform attempts haven't moved the exoneration rate, the burden is on reformers to explain why the next fifty years will be different — not on abolitionists to keep waiting while the state runs an irreversible experiment.
Jury retreat as democratic signal
C
Juries — actual citizens deliberating under oath, hearing the specific evidence about a specific person — imposed only 23 death sentences in 2025, and only 15 of more than 50 capital juries unanimously agreed to impose death. That is not a cultural trend. That is the jury system working exactly as designed, and the people closest to the facts are retreating from capital punishment faster than any elected official.
L
The conservative framing actually understates this. You have an executive branch — via the Bondi directive ordering prosecutors to seek death 'whenever reasonable' — administratively overriding the very democratic institution conservatives claim to trust. The jury is the democratic check, and the executive is systematically working around it. That's not a legitimacy gap, it's a legitimacy crisis.
C
The Bondi directive is genuinely alarming, and conservatives who care about ordered liberty should say so plainly — 'whenever reasonable' inverts the gravity the punishment requires. But that is an argument against Bondi's directive, not against capital punishment in principle.
L
A punishment that requires executive overreach to sustain itself against historic jury resistance has lost its democratic foundation — that argument cuts against the practice, not just against one memo.
Racial bias as structural, not incidental
C
The 41% Black composition of death row and the majority-Black composition of exonerees is a serious indictment — but the conservative response is not to defend the existing system. It is to demand prosecutorial accountability, rigorous evidentiary standards, and the structural scrutiny that limited-government conservatism has always applied to state power. The answer to government incompetence is sometimes discipline, not abolition.
L
The conservative prescription — more scrutiny, higher standards, prosecutorial accountability — has been the reform agenda for fifty years. What it has produced is a system where more than half of 200 exonerees are Black, and where the Bondi directive now instructs prosecutors to seek death 'whenever reasonable.' The structure is not failing to apply the right standard; it is functioning as built.
C
Saying the system is 'functioning as built' is a claim about intent that the data supports in outcomes but doesn't prove in design — and that distinction matters for what remedy is appropriate. But I'll concede this: if the reform argument requires assuming good faith from the same institutional actors producing these outcomes, that assumption is in serious trouble.
L
A concession that the good-faith assumption is 'in serious trouble' is actually the abolitionist argument — you cannot reform a system you cannot trust, and you cannot trust a system with this record operating an irreversible punishment.
Retributive case for extreme guilt
C
A conservatism that takes moral proportionality seriously cannot simply wave away the retributive argument. Timothy McVeigh killed 168 people, guilt established beyond any conceivable doubt — the state's retributive function in that case seems not just defensible but required. Burke and Scalia both understood that proportionate punishment is a legitimate expression of a moral community's values, not mere vengeance.
L
McVeigh is the case abolitionists are supposed to lose, and even there the argument is thinner than it looks: life without parole removes him from society permanently, which achieves every practical goal execution does. What execution adds is the retributive satisfaction of the killing itself — and asking whether the state should be in the business of providing that satisfaction is exactly the right question when the same system executes Cameron Todd Willingham.
C
Life without parole is not an equivalent answer to 168 dead children — the conservative tradition has never been purely utilitarian, and 'practical goals' is not the only register in which punishment operates. The moral weight of proportionate response to mass murder is real, not reducible to incapacitation.
L
If the retributive argument requires trusting the state to identify the right people with the certainty McVeigh represents, the exoneration rate tells you how often it actually achieves that — and the answer is not often enough to justify an irreversible punishment.
Majority support versus democratic legitimacy
C
The 52% approval figure is the lowest since 1972, but majority support is still majority support, and abolitionists have lost direct ballot fights in most states. The democratic legitimacy argument cuts both ways — you cannot simultaneously cite jury behavior as proof of democratic withdrawal and dismiss the polling that still shows majority approval.
L
The conservative rebuttal makes this point itself: the 52% measures abstract approval of capital punishment as a concept, not willingness to sentence a specific person to death after hearing specific evidence. Those are different democratic inputs, and when they conflict — 52% support in the abstract, 15 unanimous verdicts out of 50-plus trials in practice — the jury data is the more meaningful signal because it requires actually doing the thing.
C
That distinction is genuinely compelling — abstract poll support and concrete jury willingness are measuring different things, and I'll grant the jury signal is harder to dismiss. But democratic legitimacy also includes legislative majorities that have not moved to abolition, and that institutional inertia is not nothing.
L
Legislative inertia in the face of a declining jury mandate and a declining poll mandate isn't democratic legitimacy — it's institutional lag, and it's exactly the kind of gap that produced Furman in 1972 the last time support hit these levels.
Conservative's hardest question
The retributive argument for capital punishment in cases of absolute certainty — a McVeigh, a Boston Marathon bomber — remains genuinely powerful, and a conservatism that takes moral proportionality seriously cannot simply wave it away. If I argue the system should be reformed rather than abolished, I must answer why reform has failed for fifty years without producing the precision the argument requires.
Liberal's hardest question
The 52% public support figure — while the lowest since 1972 — still represents majority approval, which complicates the democratic legitimacy argument against execution. If the claim is that juries and public opinion are withdrawing consent from capital punishment, a skeptic can fairly note that majority support, however thin, is still majority support, and abolitionists have not yet won that argument at the ballot box in most states.
Both sides agree: Both sides accept the same core statistic — one exoneration for every 8.3 executions — as legitimate and dispositive evidence that the system as currently administered is broken.
The real conflict: A genuine factual and causal dispute: the liberal argues racial disparity in death row composition and exoneration rates reflects systemic bias embedded in institutional design, while the conservative attributes disparity to prosecutorial discretion and crime-rate differences — both acknowledge the statistics, but contest what produces them.
What nobody has answered: If the jury system is the democratic institution both sides trust most — and juries are declining to impose death at historic rates — what is the legitimate basis for an executive branch to expand capital prosecutions against that signal, and does any answer to that question leave executive power in criminal justice intact?
Sources
- Death Penalty Information Center — 2025 execution statistics and death row population data
- Gallup — October 2025 poll on death penalty support by party, age, and year-over-year trends
- Executive Order 14164, signed January 20, 2025
- AG Pam Bondi memorandum to DOJ prosecutors, February 5, 2025
- Supreme Court docket records — denial of execution stays in 2025
- Kennedy v. Louisiana, 554 U.S. 407 (2008)
- Furman v. Georgia, 408 U.S. 238 (1972)
- Gregg v. Georgia, 428 U.S. 153 (1976)