1. Introduction
People v. Shove (Cal. Aug. 13, 2026) is an automatic appeal from a Los Angeles County judgment of death.
The prosecution’s theory was that Theodore Churchill Shove III masterminded a multi-part scheme to obtain the Souther
family’s aerospace surplus business, Cal Aero: he recruited a friend (Lewis Hardin) to murder elderly owners Hubert and
Elizabeth Souther, arranged a burglary of Cal Aero’s safe, and then attempted to extort the Southers’ children using a false identity.
The case raised numerous issues typical of complex capital litigation: suppression of digital evidence seized from a home computer,
exclusion of third party culpability theories, admissibility of inflammatory recordings and “mafia” alias evidence, limits on closing argument,
victim impact evidence, and—most consequentially for future jury-selection litigation—how appellate courts should review a
Batson v. Kentucky/People v. Wheeler claim where the prosecutor offered reasons for some strikes before the trial court ruled,
but gave no reason for one strike.
The majority (Kruger, J.) affirmed the judgment in full; Evans, J., dissented (joined by Liu, J.), arguing the Batson/Wheeler handling
required reversal in light of Pitchford v. Cain (2026) ___ U.S. ___ [146 S.Ct. 1345] (Pitchford).
3. Analysis
A. Precedents Cited
1. Search warrants, probable cause, nexus, and good-faith reliance
The court’s suppression analysis is anchored in the modern “practical, commonsense” probable-cause framework of
Illinois v. Gates (1983) 462 U.S. 213 (Gates), as adopted and reiterated in
People v. Beck and Cruz (2019) 8 Cal.5th 548,
People v. Westerfield (2019) 6 Cal.5th 632, and
People v. Carrington (2009) 47 Cal.4th 145 (Carrington).
The key doctrinal move, however, is the opinion’s decision to bypass the merits of probable cause and decide on
Leon grounds, applying California’s articulation in
People v. Camarella (1991) 54 Cal.3d 592 (Camarella).
The opinion is careful to emphasize the “nexus” requirement—probable cause to believe evidence will be found in the place searched—citing
People v. Frank (1985) 38 Cal.3d 711 and People v. Gonzalez (1990) 51 Cal.3d 1179.
But it relies on cases allowing magistrates and officers to draw inferences about where evidence is likely kept:
People v. Sandlin (1991) 230 Cal.App.3d 1310,
U.S. v. Fannin (9th Cir. 1987) 817 F.2d 1379,
U.S. v. Kvashuk (9th Cir. 2022) 29 F.4th 1077 (Kvashuk),
People v. Lazarus (2015) 238 Cal.App.4th 734, and
People v. Cleland (1990) 225 Cal.App.3d 388.
For digital extortion evidence, it draws especially on
U.S. v. Adjani (9th Cir. 2006) 452 F.3d 1140 (computer-related extortion supports home-computer search).
On staleness, the court cites U.S. v. Hay (9th Cir. 2000) 231 F.3d 630 and Kvashuk for the proposition that
computer evidence can persist even after attempted deletion—supporting an officer’s reasonable belief despite a multi-month delay.
2. Third party culpability
The court reaffirms Hall as the governing admissibility framework: third party evidence must do more than show
motive or opportunity; it must link the third party to actual perpetration. The opinion defends Hall against constitutional attack,
distinguishing Holmes v. South Carolina (2006) 547 U.S. 319 and rejecting reliance on State v. Meister (2009) 148 Idaho 236.
The court situates Hall within consistent California applications: People v. Bracamontes (2022) 12 Cal.5th 977,
People v. Dworak (2021) 11 Cal.5th 881 (Dworak),
People v. Turner (2020) 10 Cal.5th 786 (Turner), and others.
3. Character/intimidation evidence: recording and alias
For the “generic threat” recording, the court’s analysis is shaped by
People v. Karis (1988) 46 Cal.3d 612 (Karis) and People v. Rodriguez (1986) 42 Cal.3d 730,
focusing on remoteness, vagueness, and whether the victim is within the threat’s scope.
Prejudice/harmlessness is assessed under People v. Watson (1956) 46 Cal.2d 818 (Watson) and cases like
People v. Chism (2014) 58 Cal.4th 1266 (Chism) and People v. Riccardi (2012) 54 Cal.4th 758 (Riccardi).
For the “Tony Bonanno” alias evidence, the court leans on relevance/undue prejudice principles and analogous nickname decisions:
People v. Lee (2011) 51 Cal.4th 620 (Lee), People v. Brown (2003) 31 Cal.4th 518 (Brown),
and intimidation/mystique evidence cases such as People v. Allen (1986) 42 Cal.3d 1222.
It also invokes the modern framing of “prejudice” under Evidence Code section 352 via People v. Nadey (2024) 16 Cal.5th 102 (Nadey).
4. Batson/Wheeler procedural posture and comparative analysis
The opinion’s most practically significant discussion concerns the “stage” of review when the trial court hears reasons before finishing step one.
It draws on:
People v. Scott (2015) 61 Cal.4th 363 (Scott),
People v. Krebs (2019) 8 Cal.5th 265 (Krebs),
People v. Hardy (2018) 5 Cal.5th 56 (Hardy),
Chism,
Riccardi,
and “hybrid” cases like People v. Mills (2010) 48 Cal.4th 158 (Mills).
On first-stage “substantial evidence” review, it cites People v. Battle (2021) 11 Cal.5th 749 (Battle).
For comparative juror analysis and deference, it references People v. Lenix (2008) 44 Cal.4th 602,
People v. Armstrong (2019) 6 Cal.5th 735 (Armstrong), and multiple post-Lenix decisions.
It also addresses (in footnote) Pitchford, distinguishing it and declining to treat it as undermining California’s understanding that a
“race neutral” statement can function as a third-stage determination in context.
5. Closing argument limits and victim impact
The court’s Griffin discussion relies on the “comment on state of evidence” line from cases like
People v. Castaneda (2011) 51 Cal.4th 1292 (Castaneda), Valdez, and related authority.
On sympathy appeals, it applies People v. Steskal (2021) 11 Cal.5th 332 (Steskal),
People v. Amezcua and Flores (2019) 6 Cal.5th 886 (Amezcua),
People v. Stansbury (1993) 4 Cal.4th 1017 (Stansbury),
and People v. Kipp (2001) 26 Cal.4th 1100 (Kipp).
For victim impact evidence, the court follows Payne v. Tennessee (1991) 501 U.S. 808 (Payne) and California’s adoption in
People v. Edwards (1991) 54 Cal.3d 787, while noting Bosse v. Oklahoma (2016) 580 U.S. 1 and
Booth v. Maryland (1987) 482 U.S. 496 do not bar the evidence presented here.
B. Legal Reasoning
1. The “good faith” bridge from weak nexus to admissibility in digital-evidence searches
The court’s suppression holding reflects an increasingly pragmatic judicial posture toward digital evidence.
The affidavit did not explicitly say evidence would be in the residence, and the defense argued both “no nexus” and “staleness.”
The court nonetheless held that, even if probable cause were contestable, Leon saved the search because the affidavit was not
“so lacking in indicia of probable cause” that reliance was unreasonable.
The reasoning turns on two inferences officers could reasonably make:
(1) extortion instruments were computer-generated; and
(2) it is normal to infer that computers and stored files are likely found in the suspect’s home.
The staleness concern was neutralized by the nature of digital evidence (persistence/recoverability).
The case thus reinforces a pattern: in close digital-nexus cases, courts may prefer a Leon resolution to avoid brittle, formalist nexus rules.
2. Hall reaffirmed: “motive/opportunity” is not a defense strategy by itself
The opinion reads Hall not as a special, heightened barrier but as ordinary relevance doctrine:
evidence that does not create a reasoned link to perpetration does not “raise a reasonable doubt” and so is not relevant.
The court then applies that filter to a sprawling defense narrative about Kenneth (inheritance, demeanor, footprints, phone box, cash sightings)
and finds it either (a) motive/opportunity alone; (b) speculation; or (c) collateral impeachment.
A key structural point is the court’s insistence on “specificity”: third party evidence must connect a particular person to the crime,
not simply raise the possibility that “someone else” could have done it (citing Turner).
This is especially important in multi-perpetrator theories: even if a third party were involved, that would not necessarily exculpate Shove
given the prosecution’s “mastermind” theory.
3. Evidence of persona and intimidation: alias admissible; remote “killer” recording not
The court separates two kinds of persona evidence:
-
Alias/mafia-posturing: relevant to intent and method—how Shove pressured, persuaded, and controlled others in the scheme.
The prejudice risk (jurors think “mafia = criminal”) was mitigated by evidence that the “connections” were fantasy and by the prosecutor’s framing.
-
Old recorded monologue about killing: too remote, vague, and unmoored from the charged crimes to show state of mind at the time;
risked impermissible propensity; therefore, admission was error, though harmless given the volume of other evidence.
4. The opinion’s novel procedural move: splitting Batson/Wheeler review by juror
The majority treats the Batson/Wheeler ruling as a hybrid: where reasons were heard before the prima facie ruling,
it proceeds to third-stage review for jurors who were actually explained. But it then introduces a “wrinkle”:
because the prosecutor never offered a reason for Prospective Juror No. 206, the court reviews that strike as a first-stage denial.
Substantively, this creates a juror-by-juror staging method:
step three for strikes with articulated reasons; step one for the unexplained strike, even though the motion was global.
The court justifies this as necessary because the “third stage” could not have addressed reasons never given.
The dissent rejects this bifurcation as inconsistent with Scott’s emphasis on predictable procedure and with the
practical reality that the trial court’s “race-neutral” comment did not reflect a completed step-three genuineness assessment,
especially after Pitchford.
5. Trial fairness: policing emotion without rewriting trials
The court candidly finds some guilt-phase sympathy argument improper, but applies a familiar harm analysis:
brevity, lack of repetition, and curative instructions (“must not be influenced by pity”) can render such error nonprejudicial.
This reflects a consistent line in California capital review: courts will identify overreach but resist reversal absent
a plausible pathway from the rhetoric to the verdict in a record with strong evidence.
C. Impact
1. Batson/Wheeler litigation: a new fault line—“unexplained strike” and appellate staging
The most likely lasting effect of People v. Shove lies in how it manages an incomplete record:
when the prosecutor provides reasons for some challenged strikes but omits one, the majority permits appellate courts to
(a) conduct third-stage review for the explained strikes while (b) preserving a first-stage analysis for the unexplained one.
Practically, this incentivizes defense counsel to immediately insist on a reason for each targeted strike once the prosecutor begins offering reasons,
and to request an explicit step-three finding on genuineness—especially after Pitchford spotlighted step-three breakdowns.
The dissent signals that federal doctrine may be moving toward stricter insistence that defendants be afforded a meaningful opportunity
to argue pretext and that trial courts must actually decide it.
2. Digital evidence searches: Leon as the stabilizer
On suppression, the opinion reinforces that when crimes involve computers (drafting, printing, storing communications),
courts may accept home-search inferences as reasonable enough to satisfy Leon even where the affidavit is thin on explicit residential nexus.
This matters for extortion, harassment, and fraud cases where the core instrumentality is a document created electronically.
3. Defense strategy: third party culpability must be built like a case, not a narrative
The opinion reiterates that capital defendants cannot use Hall to introduce broad suspicion of family members or insiders.
Unless proffers concretely connect a named third party to perpetration, courts may exclude them as speculative and distracting.
In practice, this raises the premium on developing admissible linkage evidence early (physical evidence, communications, admissions)
rather than relying on “motive plus odd behavior.”