Callaham v. United States, 268 A.3d 833

Background

Defendant Timothy Callaham was tried for robbery based primarily on about 7.5 minutes of silent surveillance footage; the alleged victim did not testify.
Two detectives authenticated and played compiled camera clips and gave testimony pointing out events on the video; the prosecutor prepared the compiled clips.
During deliberations jurors sent multiple substantive notes asking whether the footage established a robbery and twice reported deadlock.
The trial court first gave a Winters anti-deadlock instruction, the jury then announced a verdict that broke down during polling, and the court thereafter gave a Crowder-style poll-breakdown instruction containing language similar to Winters.
After the Crowder instruction the jury returned a compromise verdict convicting Callaham of the lesser included robbery.
The D.C. Court of Appeals reversed, holding there was a substantial risk the verdict was coerced; the court also addressed confrontation, best-evidence, and lay-witness narration claims and mostly rejected them while reaffirming personal-knowledge limits for lay testimony.

Issues

Issue	Plaintiff's Argument (Callaham)	Defendant's Argument (Government)	Held
Jury coercion from post-poll Crowder instruction	Crowder instruction (with Crowder bracketed language) coerced jurors after Winters instruction and a poll breakdown	The Crowder instruction was appropriate to resolve a revealed lack of unanimity and not coercive	Reversed: substantial risk of coercion; Crowder instruction functioned as a second anti-deadlock charge and risked coerced verdict
Sixth Amendment Confrontation Clause — failure to call prosecutor who compiled video	Excluding the prosecutor who compiled the clips denied Callaham the right to confront a testimonial witness who ‘created’ the exhibit	Compiling sequential clips from authenticated footage is not a testimonial statement and did not create new testimonial evidence	Denied: no Confrontation Clause violation; compiler did not make testimonial statements for prosecution
Best evidence rule — detective testimony about video without original footage	Detective Volpe’s testimony describing video content violated the best evidence rule because originals were not introduced	The government did introduce authenticated compiled clips (duplicates); detectives’ testimony did not supplant originals	Denied: best-evidence claim fails because admitted compiled video was before jury and duplicates are admissible
Lay-witness narration of surveillance video by detectives	Detectives improperly narrated events they did not personally observe in real time	Government argued examining and describing video is like identifying photograph; narration aided jury orientation	Court reaffirmed lay-witness personal-knowledge requirement, criticized narration but endorsed limiting instructions; did not find reversible error on this record given possible changes at retrial

Key Cases Cited

Crowder v. United States, 383 A.2d 336 (D.C. 1978) (poll-breakdown instruction language and caution about juror coercion)
Winters v. United States, 317 A.2d 530 (D.C. 1974) (classic anti-deadlock instruction and explanation of its coercive potential)
Green v. United States, 740 A.2d 21 (D.C. 1999) (jury poll purpose and protecting unanimity against coercion)
Davis v. United States, 669 A.2d 680 (D.C. 1995) (probabilistic standard for assessing substantial risk of coerced verdict)
Coley v. United States, 196 A.3d 414 (D.C. 2018) (appellate standard assessing risk of coercion)
Hankins v. United States, 3 A.3d 356 (D.C. 2010) (evaluate coercion from jurors' perspective and judge’s responses)
Fortune v. United States, 65 A.3d 75 (D.C. 2013) (reversing where repeated deadlock and pressure elevated coercion risk)
United States v. Smith, 640 F.3d 358 (D.C. Cir. 2011) (discussed for comparison on creation of testimonial records)
Holmes v. United States, 92 A.3d 328 (D.C. 2014) (surveillance system as a tool; does not eliminate personal-knowledge rule for lay witnesses)