The Scipio Files · argument map
Fifteen documents, and unlike most topics in this series, the Supreme Court argues with itself directly across five cases, quoting and rebutting its own past opinions by name. Ehrlich's econometrics, Sunstein and Vermeule's moral argument, and the Criminal Justice Legal Foundation's Kent Scheidegger make the retentionist case. Marshall, Brennan, Souter, Breyer, Carol Steiker, and the Death Penalty Information Center make the abolitionist case. The National Academies, the GAO, the Bureau of Justice Statistics, and the Urban Institute supply the numbers both sides invoke. Hover a card to see the assumption underneath it, click it for the full exchange, and ask which claims are actually arguing with each other and which are simply describing different things.
Color marks the side of the volley, not who is right. K marks a source that is neutral because it is a single official, congressional, or nonpartisan-research estimate both sides cite selectively (the National Academies' deterrence review, the GAO's racial-disparity synthesis, BJS's capital punishment census, and the Urban Institute's state-commissioned cost study). Justices, scholars, and advocacy organizations keep the same M or D label in every card regardless of whether that specific card's finding cuts toward or away from the death penalty that day, the same source-identity discipline 10k applied to CBO and the Commonwealth Fund. One judgment call worth double-checking: Justice Stewart is coded M throughout even though his own Furman concurrence supplies the dataset's most quoted abolitionist-sounding line (the 'struck by lightning' passage in A1), because Stewart never held the death penalty itself unconstitutional, only its arbitrary application, and he went on to co-author the Gregg plurality four years later upholding a reformed statute on that basis.
A caution before you count silences. The retentionist side holds 7 of 14 last words (50 percent) on 19 percent of the dataset's 36 cards. That concentration traces almost entirely to two sources: Justice Scalia's 2015 Glossip concurrence supplies 3 of the 7 (on delay and decline, deterrence, and retribution), and Kent Scheidegger's 2011 report to the Connecticut legislature supplies 3 more (on race-of-victim litigation, the practical rarity of wrongful convictions, and the length of capital review). Lewis Powell's McCleskey majority supplies the seventh (the slippery-slope objection to statistical claims). It reflects which two documents happened to be written as direct, multi-issue rebuttals rather than an even distribution of argumentative strength: Scalia's concurrence exists specifically to answer Breyer's dissent point by point in the same case, and Scheidegger's report exists specifically to rebut cost and disparity claims made to a state legislature. The abolitionist side's 6 last words (B3, B6, D6, E3, E5, F5) and the single neutral last word (A6, the Bureau of Justice Statistics) are spread across different documents and years. Separately: two of Scalia's own citations for a deterrent effect (Zimmerman 2004 and Dezhbakhsh, Rubin, and Shepherd 2003) are not independently on file in this dataset, an honest limitation flagged on card C6 itself. On cost specifically: Lane F's abolitionist voice (F5, Carol Steiker, 2005) makes a general opportunity-cost argument, that money spent on any execution could save more lives if spent on ordinary public purposes instead, published three years before the Urban Institute's Maryland-specific cost study (F1 and F2, 2008) and six years before Scheidegger's reply to it (F3 and F4, 2011). It does not engage either document's own dollar figures directly, so read it as the abolitionist side's own parallel case for why cost cuts against retention, not as a rebuttal to Scheidegger's specific 'cost mirage' argument.
| Card | Question | Speaker | Answers | Move | Claim | Cite | Last word |
|---|