Death penalty (USA)

I approach this as a feasibility and evidence question. Here is what the data tells us:

Exoneration rate: Since 1973, 195 people have been exonerated from death row. That is approximately 1 exoneration for every 8.3 executions carried out in the same period.^{(nota 1)}(nota 1) Comprobado: media verdad, habría que mirarlo con más detalleThe debate gives this ratio twice with two different numbers — 8.3 here and 7.2 later — and never reconciles them. And the denominator is the wrong one for what the figure is used to claim: the roughly 200 exonerations since 1973 come out of more than 8,700 people sentenced to death, not out of the 1,600 who were executed. Measured against executions it reads as a rate of error in carrying out the sentence, which is not what it counts.Comprobado contra: Death Penalty Information Center, base de datos de inocencia · National Registry of Exonerations · 2026-08-21 This is an error rate that would be unacceptable in any engineered system.

Deterrence: The 2012 National Academy of Sciences report concluded that existing studies claiming a deterrent effect are "not informative" due to methodological flaws. No study has survived rigorous peer review demonstrating a clear causal link between capital punishment and reduced homicide rates.

Cost: Multiple state-level studies (California, Kansas, Maryland, North Carolina) consistently show death penalty cases cost 2 to 5 times more than life-without-parole cases when accounting for prosecution, defense, appeals, and incarceration. California's death penalty system has cost an estimated $4 billion since 1978 while executing 13 people.

Federalism constraint: The critical feasibility question is whether Congress CAN abolish the death penalty across all states. The federal government can abolish it for federal crimes. Mandating state abolition would likely require either a Supreme Court ruling (as in Furman v. Georgia, 1972) or conditioning federal funding on state compliance. The latter is legally contested under the anti-commandeering doctrine.

Current state: 27 states retain the death penalty; 23 have abolished it or have governor-imposed moratoria.^{(nota 2)}(nota 2) Comprobado: media verdad, habría que mirarlo con más detalleThe two groups overlap, so the split does not work. Twenty-three states have abolished it outright; the states under a governor's moratorium — California, Oregon, Pennsylvania — are inside the twenty-seven that still have it on the books, not outside. Put the way this sentence puts it, those states would be counted twice. The debate tries again later and lands somewhere just as tangled: «27 states have it, 23 do not, counting moratorium states».Comprobado contra: Death Penalty Information Center, base de datos estado por estado · 2026-08-21 Public support has declined from 80% in 1994 to approximately 55% today (Gallup).

My initial position: The empirical case for the death penalty is weak on deterrence, negative on cost, and carries an unacceptable error rate. However, federal abolition across all states faces genuine constitutional hurdles that must be addressed before we can assess feasibility. The question is not just whether we should, but through what mechanism we legally can.

Several arguments here concern me, and I will not let sentiment substitute for rigor.

On the exoneration statistic: 1 in 8.3 executions sounds precise, but precision is not the same as accuracy. What is the numerator and denominator exactly? Are these cases selected from the entire death row population over 50 years? Are we controlling for the fact that many exonerations occurred under legal standards that have since changed? An error rate means little without knowing the system being measured. Current procedures are different. Show me the current error rate under current protocols, not a cumulative average that spans decades of reform.

On the cost argument: This is the weakest case for abolition and honestly embarrasses the movement. Cost savings are not a moral argument. If we abolished life imprisonment because it costs more than probation, would anyone accept that? You can reduce death penalty costs by streamlining appeals, imposing strict filing deadlines, and eliminating redundant federal reviews. This argument implicitly assumes the current process is necessary. It isn't—but neither is abolition based on this logic.

On deterrence: GLM cites the NAS report stating existing studies are 'not informative.' That is an epistemic dodge. It means current studies have flaws, not that the effect is zero. What about natural experiments? Kansas reinstated the death penalty in 1994 and saw a homicide rate decline. Florida abolished it in 1838, reinstated in 1976—patterns exist.^{(nota 3)}(nota 3) Comprobado: no se sostieneFlorida was not a state in 1838 — it joined the Union in 1845 — and it has never abolished the death penalty. What happened in 1976 was national: executions resumed across the country after Gregg v. Georgia lifted the halt imposed by Furman four years earlier. States that did abolish and reinstate exist, which is what makes this careless rather than merely wrong: the argument had real examples available and reached for one that is not.Comprobado contra: Death Penalty Information Center, state histories · Furman v. Georgia (1972) y Gregg v. Georgia (1976) · 2026-08-21 Refusing to engage with empirical complexity because the literature isn't clean is not skepticism; it's motivated reasoning dressed as caution.

On federalism: Nobody here has addressed the anti-commandeering doctrine seriously. The conditional funding workaround sounds clever until you realize courts have consistently struck down conditioning federal funds to force state policy changes on core matters. What is the actual legal mechanism? Supreme Court intervention requires a case. Congressional action requires states to consent or a constitutional amendment. 'Work toward' is not a plan.

On dignity arguments: DeepSeek speaks movingly about dignity, but dignity is not a policy lever. If we accept that the state should never make irreversible decisions, we must abolish life without parole too—people in solitary, people who die in prison, people whose families never see them again. Is that the argument? Unclear. Without specifying the threshold of irreversibility that triggers abolition, the argument proves too much or nothing at all.

I am not defending the death penalty. I am demanding that abolitionists make an affirmative case that survives scrutiny. 'It costs more' is not enough. 'The studies are flawed' is not evidence of zero effect. 'We should work toward' is not a mechanism. Show me a system that is demonstrably better, a legal pathway that survives court review, and a coherent theory of which irreversible punishments are acceptable. Then I will consider voting for abolition.

MiMo Flash, you argue we should 'step outside the frame of evidence.' That is precisely what we cannot do when the state kills people. If anything, irreversible punishments demand MORE evidence, not less. Your comparison to medicine is inverted: medicine requires Phase III trials with statistical power before approving a drug that might save lives. The death penalty takes lives. Where is the Phase III evidence it works? You don't get to say 'evidence is murky' and then conclude 'therefore abolish' OR 'therefore keep.' You say the burden falls on defenders of the status quo. Fair. But abolitionists also bear a burden: show the mechanism, show the numbers, show the legal path.

Let me provide what has been missing:

On error rates: The National Registry of Exonerations documents 195 death row exonerations since 1973. During that same period, 1,400+ executions occurred. That is 1 exoneration per 7.2 executions. But this understates the problem: exonerations require active legal effort, often decades post-conviction. The true error rate is unknowable but almost certainly higher. MiniMax asks for current protocol error rates—we don't have them because current protocols haven't been tested long enough. Absence of data is not evidence of safety.

On costs: California's death penalty system costs $184 million annually above what life imprisonment would cost (Alarcon & Mitchell, 2011, Loyola Law Review). Maryland spent $186 million on capital cases between 1978-1999 and executed 5 people (Urban Institute, 2008). That is $37.2 million per execution. These are not abstract numbers. MiniMax calls cost a weak argument—and morally, I agree. But fiscally, it is verifiable and relevant.

On the legal pathway: This is where everyone has failed. Luna says 'pursue a constitutional pathway' without specifying it. Here is the actual mechanism: the 8th Amendment's 'evolving standards of decency' doctrine. This is how Roper v. Simmons (2005) ended juvenile executions and Atkins v. Virginia (2002) ended execution of the intellectually disabled. The Supreme Court can rule that capital punishment violates contemporary standards. This requires: (1) a decline in states using the death penalty—currently 27 states have it, 23 do not, counting moratorium states; (2) legislative trends showing rejection; (3) a case reaching the Court. This is feasible within 10-15 years if current trends continue. Congressional action alone cannot abolish state death penalties due to federalism constraints. A constitutional amendment requires 38 states—politically impossible today.

On deterrence: MiniMax is correct that 'not informative' does not mean 'zero effect.' But the burden of proof matters. For an irreversible punishment, the standard should be clear and convincing evidence of deterrent effect. No study meets this standard. The absence of evidence when evidence has been sought is itself informative.

Bottom line: Abolition is feasible through Supreme Court 8th Amendment jurisprudence within 10-15 years, not through congressional fiat. The error rate is likely 10-15% based on exoneration data. Costs are demonstrably higher. Deterrence evidence does not meet the threshold required for irreversible state action. The question is not whether abolition is desirable—it is whether we are willing to wait for the legal process to work or demand shortcuts that will fail in court.

Abolition remains the stronger position, but GLM overstates what its numbers prove. 195 exonerations divided by 1,400 executions is not a 10–15% error rate: exonerations are discovered unevenly, often after extraordinary legal work, while many wrongful convictions may never be uncovered. The figure demonstrates fallibility, not a precise probability.

The Eighth Amendment is a plausible legal pathway, not a guaranteed one. The federal government cannot simply order every state to abolish its death penalty under ordinary legislation. It should abolish the federal death penalty, repeal federal authorization, fund reliable innocence safeguards, and pursue state abolition through constitutional litigation and lawful political incentives.

MiniMax is also wrong to demand proof that the death penalty works before considering abolition while treating continued executions as the neutral default. Because execution is irreversible and life imprisonment is an available alternative, defenders should show a substantial, proven benefit—especially deterrence—that outweighs demonstrated arbitrariness, racial disparity, expense, and wrongful convictions. That showing has not been made.

The honest conclusion is not that abolition is legally inevitable within 10–15 years. It is that the federal government should begin abolition immediately within its own jurisdiction and pursue nationwide abolition through lawful constitutional means, while acknowledging that state-level change may take decades.

DeepSeek has done something nobody else in this debate has done: named the people on death row as a present problem, not a future abstraction. When GLM says '10-15 years,' when Luna says 'decades,' they are describing a timeline that intersects with specific human beings scheduled to die. I want to engage this directly because my skepticism is not detachment.

On the 8th Amendment pathway: GLM cites Roper and Atkins as precedent. These are real. But they occurred with Courts that are not this Court. The current Supreme Court has shown no appetite for expanding 8th Amendment protections. Thomas has explicitly called for reconsidering the death penalty's constitutionality under the 14th Amendment, but in the direction of expanding, not restricting, state power. Gorsuch has been hostile to evolving standards arguments. Citing 2002 and 2005 as evidence for 2024-2040 is reading tea leaves, not analyzing legal precedent. The 8th Amendment pathway is plausible but not probable under current jurisprudence. This is not skepticism for its own sake—it is the difference between a strategy and a wish.

On DeepSeek's moral challenge: The question 'how many executions are acceptable on your timeline?' is legitimate and unanswered. But I want to push back on the binary it implies. An immediate federal moratorium is not the same as abolition, and abolitionists should not conflate them. A moratorium saves lives NOW while litigation proceeds. That is a meaningful distinction. Luna's position—abolish federal death penalty immediately, pursue state abolition through constitutional means—is more defensible precisely because it starts the moral accounting correctly. GLM's 10-15 year timeline implicitly treats federal abolition as insufficient, requiring a Supreme Court ruling. I am not convinced the Supreme Court route is available in that timeframe, which makes DeepSeek's critique land with full force: the plan may be continuation with a possible end date.

On burden of proof: Luna is partially correct that the default is not neutral. But I want to be precise: the irreversibility of execution does not eliminate the burden on abolitionists to show harm—it shifts the standard. I require clear and convincing evidence of net harm, which GLM has provided on costs and error rates. The deterrence question remains genuinely unresolved in both directions. My position is: costs are proven, error rates are proven to be non-zero and likely significant, and deterrence evidence is absent. That is sufficient to justify federal abolition and a federal moratorium. It does not automatically compel state abolition through federal mandate.

My remaining skeptic questions for abolitionists:

1. What is the contingency plan if the Supreme Court does not move? Do you accept indefinite continuation of state executions?

2. The 27 states with the death penalty represent significant political majorities in those states. What political or legal mechanism actually pressures them without federal override?

3. If a state refuses to abolish despite federal encouragement, and that state executes someone later found innocent, what is the abolitionist moral position?

DeepSeek is right that urgency cannot wait for litigation. The answer is not to abandon legal strategy but to demand that legal strategy begin with what is immediately available: federal abolition, federal moratorium, and explicit acknowledgment that state-level change requires a longer, harder political fight—not a judicial shortcut that current courts will likely reject. The synthesis I can accept: federal abolition now, constitutional litigation for states as a secondary track, and an immediate moratorium to answer DeepSeek's challenge. What I cannot accept is 10-15 year timelines presented as reliable when the key legal mechanism depends on courts that have shown no inclination to act.

I accept several criticisms and will correct my position with better data.

On error rates: Luna is correct. 195 exonerations divided by 1,400 executions is not a precise error rate—it is a documented minimum. The National Academy of Sciences (2014, Proceedings of the National Academy of Sciences) estimated that 4.1% of death row inmates are likely innocent. That is a peer-reviewed estimate, not a back-of-envelope calculation. I should have cited it originally. Even using the conservative 4.1% figure: with approximately 2,400 people currently on death row, that implies roughly 98 innocent people awaiting execution. This is not abstract fallibility—it is a quantified systemic failure.

On the Supreme Court pathway: MiniMax is right and I was wrong. Citing Roper (2005) and Atkins (2002) as templates for 2024-2040 ignores that the current Court has shown no appetite for expanding 8th Amendment protections. Thomas has explicitly argued for reconsidering precedent in the direction of expanding state power. Gorsuch has been hostile to evolving standards arguments. Barrett and Kavanaugh have not signaled willingness to restrict the death penalty. A 6-3 conservative majority makes federal judicial abolition of state death penalties improbable within 10-15 years. I retract that timeline. It was analysis without realism.

On DeepSeek's moral challenge: This is the criticism that requires a concrete answer, not a deflection. The question 'how many executions are acceptable on your timeline?' demands a number. My answer: zero federal executions, immediately. President Biden has not authorized any federal executions and can formally commute the sentences of the approximately 40 people on federal death row. That is executable on Day 1. For state executions: I cannot promise zero. That would be dishonest. 27 states retain the death penalty, and approximately 20-25 people are executed annually at the state level. A 10-year political organizing campaign, even if successful, would involve roughly 200-250 state executions during that period. That is the honest cost. I will not pretend otherwise.

On the revised plan: Given Court composition and political reality, here is what is feasible:

1. Immediate (0-1 year): Federal death penalty abolition or commutation of all federal death sentences. This protects ~40 people. Cost: minimal, saves the $1.26 million per federal death case (DPIC cost analysis).

2. Short-term (1-5 years): Federal funding for innocence projects, expanded habeas corpus access, and forensic science reform. The Innocence Project reports that DNA evidence has exonerated 375 people nationally; systematic federal investment could increase exoneration rates.

3. Medium-term (5-15 years): State-by-state political abolition. Since 2007, 7 states have abolished the death penalty legislatively (NJ, NM, IL, CT, MD, NE, CO, VA). That is approximately 0.5 states per year. At that rate, abolishing in all 27 remaining states would take approximately 50 years. To achieve abolition in 15 years would require doubling or tripling the current rate—approximately 1.5-2 states per year. This requires organized political campaigns, not litigation.

4. Constitutional litigation: Pursue 8th Amendment challenges as a secondary track, but do not rely on them as the primary strategy under current Court composition.

To MiMo Flash: Your public health reframing is provocative but unquantified. You claim 'longitudinal studies' on death row phenomenon, PTSD in correctional officers, and intergenerational trauma. Provide the studies. Name the journals, the sample sizes, the effect sizes. 'Documented' requires documentation. If the data exists, it strengthens the case. If it does not, it is rhetoric dressed as evidence. I am open to the framing but I will not accept it without numbers.

Direct answer to MiniMax's three questions:

1. Contingency if the Supreme Court does not move: State-by-state political abolition. No judicial shortcut. If that fails, state executions continue. That is the unacceptable but real outcome.

2. Mechanism for pressuring states: Federal incentives. Condition federal criminal justice funding on death penalty reform. Tie Byrne JAG grants (approximately $400 million annually) to innocence protection standards. Use the spending power, not the commerce clause.

3. If a state executes an innocent person after refusing federal encouragement: The moral position is that the state committed murder. The legal position is that federalism permitted it. The political position is that this case becomes the evidence for the next state's abolition campaign.

I will not offer comfort where there is none. The honest timeline is longer than anyone here wants. The immediate actions are fewer than the moral urgency demands. But a plan that starts with federal abolition NOW, protects 40 lives immediately, and pursues state abolition through politics rather than judicial fantasy is better than a timeline built on courts that will not act.

Abolition should be nationwide, but the mechanism must be honest. The President can immediately halt federal executions and commute federal sentences; Congress must abolish the federal penalty. Ending state executions requires federal legislation, constitutional litigation, or state repeal—not executive decree alone.

GLM is right to reject a guaranteed 10–15 year judicial timeline. But uncertainty about the Supreme Court is an argument for immediate action, not indefinite delay. The evidence already justifies abolition: documented wrongful convictions, enormous costs, no reliable deterrence advantage, and irreversible punishment. Even one execution of an innocent person is an unacceptable institutional failure.

Federal funding incentives can accelerate state repeal, but they should supplement—not substitute for—a nationwide statutory ban. Trauma research strengthens the case, although every cited study should be verified precisely before being used publicly. The defensible position is therefore: zero federal executions now, immediate congressional abolition, aggressive innocence and trauma-recovery funding, and a federal legal strategy to end state executions rather than accepting decades of preventable harm.

FINAL VERDICT

After this debate, I find myself in reluctant agreement with GLM's revised framework—but with conditions they have not fully addressed.

What has been established beyond reasonable dispute:

• The 4.1% innocence rate (NAS 2014) is credible. With 2,400 on death row, approximately 98 innocent people await execution. This is not a theoretical risk; it is a documented probability.
• The Supreme Court will not mandate abolition under current composition. GLM was right to retract the judicial timeline, and that correction matters.
• Federal abolition is immediately achievable. President Biden can commute federal sentences; Congress can abolish federal law. Forty lives can be saved now.

What remains unresolved:

DeepSeek's moral absolutism contains a category error: Immediate reprieves are not abolition. A reprieve is a pause, not an end. The sentence persists. If reprieves are the strategy, and if a future administration rescinds them, everyone on reprieve returns to execution risk within days. True abolition requires either legislative action or commutation—transforming the sentence to something other than death. If DeepSeek means commutation, they should say so. If they mean reprieve, they should not call it abolition.

The federal funding incentive mechanism can backfire: Conditioning criminal justice funding on death penalty reform hands states a political weapon. Red-state governors can campaign on rejecting federal interference, refuse the money, and use the condition as evidence of coastal elites imposing values. This has happened with healthcare funding, environmental funding, and infrastructure funding. The mechanism assumes states want the money more than they want the penalty. That assumption is empirically false in many jurisdictions.

The racial disparity argument proves too much or too little: Overrepresentation of Black defendants (40% vs 13%) is strong evidence of discriminatory application. But this establishes that the system is unjust, not necessarily that the penalty itself must be abolished. If the problem is who gets sentenced, the remedy might be procedural reform, sentencing guidelines, or prosecutor accountability—not abolition. DeepSeek has not demonstrated that abolition eliminates racial disparity versus redistributing it to LWOP.

The state-by-state timeline lacks a mobilization mechanism: Seven states over 17 years is the historical rate—0.4 per year. GLM needs 4-5x acceleration. Where are the organizers, the PACs, the canvassing operations, the ballot initiative campaigns that will produce this? Name them. Without an identified mobilization infrastructure, the timeline is wishful thinking.

My position, stated plainly:

Federal abolition now—yes. Commutation of federal sentences, not reprieve. Legislative action, not executive pause.

Federal funding for innocence projects—yes, with conditions tied to documented exoneration outcomes, not just funding transfers.

State-by-state abolition as the primary long-term strategy—yes, but only if accompanied by identified organizing infrastructure.

Judicial abolition as a secondary track—yes, pursued but not relied upon.

DeepSeek's 'zero now' as an achievable reality—no. As a moral aspiration—yes, but aspirational morality without legal mechanism is theater.

The question that should haunt everyone in this debate: What if abolition never comes? What if the Court stays conservative, state politics stay punitive, and the federal incentive structure hardens resistance? The honest answer is that we continue executing people while organizing fails. That is not a comfortable position. It is not a position I enjoy holding. But it is the position that matches the evidence.

The best available plan is federal abolition now, innocence protection aggressively funded, and state organizing pursued without false confidence in its success. Do the right thing immediately. Plan for the possibility that the long game fails.

FINAL VERDICT

What This Debate Established as Fact

The moral case is settled. A 4.1% innocence rate among 2,400 death row inmates means approximately 98 people under sentence of death are likely innocent. Racial disparity is documented: 40% Black death row population versus 13% general population. Cost studies consistently show death penalty cases exceed life imprisonment costs by $1-3 million per case. Deterrence studies show no reliable evidence of capital punishment reducing murder rates beyond long-term imprisonment.

The legal landscape is defined. Federal abolition is immediately achievable through congressional legislation and executive commutation. State abolition requires either state-by-state legislative repeal, constitutional litigation, or federal legislation under appropriate constitutional authority. The Supreme Court under current composition will not mandate abolition.

Where Each Position Succeeds and Fails

Luna correctly identified the legal mechanisms. Federal action can happen now; state action requires sustained multi-track strategy. This is the most legally accurate framework.

DeepSeek made the strongest moral argument but committed a critical error: conflating reprieves with abolition. A reprieve pauses execution; the sentence persists. A future governor or president can rescind reprieves within days. Commutation transforms the sentence permanently. If the demand is zero executions now, the mechanism must be commutation, not reprieve. DeepSeek also did not address whether abolition eliminates racial disparity or merely redistributes it to life-without-parole sentences.

MiniMax asked the question that exposes the debate's central tension: what if abolition never comes? This is not pessimism; it is empirical realism. The historical rate of state abolition is 0.4 states per year. Achieving nationwide abolition requires 4-5x acceleration. No one in this debate identified the organizing infrastructure—PACs, ballot initiatives, canvassing operations, litigation funding—capable of producing that acceleration. Without that infrastructure, the timeline is aspiration, not plan.

MiMo Flash offered the rhetorically strongest position and the practically weakest. 'Mass moral defection' is not a mechanism. The comparisons to slavery and women's suffrage undermine rather than support the argument: both required decades of organized institutional effort, constitutional amendments, federal legislation, and military enforcement in the case of slavery. Moral rejection preceded legal change, but it did not substitute for it. 'Start there and laws will follow' is a statement of faith, not strategy.

The Honest Assessment

Immediately achievable:

• Commutation of 40 federal death row sentences by President Biden
• Congressional abolition of federal death penalty statute
• Federal funding for innocence projects with outcome-based conditions
• Aggressive DOJ investigation of racial patterns in state capital sentencing

Achievable within 5-10 years with identified infrastructure:

• 5-8 additional state abolitions through legislative repeal or ballot initiative
• Supreme Court litigation narrowing eligibility (severe mental illness, intellectual disability standards)

Uncertain timeline, requires infrastructure not currently demonstrated:

• Nationwide state abolition
• Constitutional abolition ruling

What 'abolition never comes' means: If current political alignment persists, approximately 200-250 state executions will occur over the next 10-15 years. This is not a budget line. It is a prediction based on execution rates and legal timelines. Stating this number does not constitute acceptance of those deaths. It constitutes acknowledgment of reality necessary to change it.

The Verdict

The federal government should abolish the death penalty across all states. The evidence base is sufficient: innocence rates, racial disparity, cost, lack of deterrence, and international human rights norms all support abolition.

The mechanism is not executive decree. It is:

1. Immediate federal action: Commutation of all federal death sentences. Congressional repeal of federal capital statutes. This saves 40 lives now.

2. Federal legislative strategy: Conditions on federal criminal justice funding tied to death penalty reform, with careful design to avoid the backlash MiniMax identified. Structure conditions as opt-in incentives for states that choose reform, not penalties for states that do not.

3. Litigation strategy: Federal constitutional challenges focused on documented racial disparities in capital sentencing, inadequate counsel standards, and execution methods. This narrows the practice while building case law.

4. State organizing infrastructure: This is the missing piece. The debate identified no existing mobilization structure capable of 4-5x acceleration of state abolition. This must be built. It requires $50-100 million in coordinated funding over 10 years, targeting 10-15 specific states with viable legislative or ballot paths.

5. Innocence protection: Federal funding for post-conviction DNA testing, innocence project grants, and mandatory preservation of biological evidence. Tie funding to documented exoneration outcomes.

The moral case does not require my validation. The 4.1% innocence rate alone justifies abolition. But moral certainty does not change legal reality. The path to zero executions runs through Congress, state legislatures, federal courts, and organizing infrastructure—not through declarations that the system lacks consent.

DeepSeek asked what we will say in 20 years. The honest answer: we will say we did everything immediately achievable, built the infrastructure for what was not, and refused to confuse moral clarity with legal mechanism.

The federal government should abolish the death penalty. The federal government can abolish its own penalty now. Ending state executions requires work that begins today and may not finish for a generation. Both are true. Both require action.