Date: Aug 1, 2026

Prospective Supervisory Reform Requires Only “Prudence and Good Sense”: A Call to Abolish Consciousness-of-Guilt Instructions

I. Introduction

In State v. Kim (Concurrence in Part & Dissent in Part) (Conn. July 28, 2026), Justice D’Auria (joined by Justice McDonald) agrees with the majority that the trial court abused its discretion by giving a “consciousness of guilt” jury instruction based on the defendant’s alleged inconsistent statements. He parts company with the majority on two fronts: (1) whether institutional constraints (supervisory-authority “limits” and stare decisis) should deter Connecticut from discontinuing the instruction, and (2) whether the instructional error was harmless.

The case arises from accusations of sexual and physical assault. The defendant, Yong Sik Kim, testified and offered an alternative narrative: the complainant initiated contact; he pushed her away and raised a hand as if to strike her; and he later kept quiet because he had promised not to tell. The state, by contrast, relied on the complainant’s account, corroborating medical testimony, photographs, and DNA evidence. The trial court additionally instructed jurors that alleged post-incident conduct—here, an asserted “change” in the defendant’s story to workplace employees—could indicate a “guilty conscience.”

Justice D’Auria’s opinion treats the consciousness-of-guilt instruction as an increasingly unjustified judicial “comment” that risks tipping credibility contests, and he argues the instruction should be abolished altogether (or, at minimum, limited to defense-requested or unobjected-to circumstances with a balanced, neutral charge).

II. Summary of the Opinion

Agreement with the majority: The trial court’s consciousness-of-guilt instruction was an abuse of discretion because it unduly magnified weak evidence and adopted a pejorative framing (“change[d] his story”) aligned with the state.
Disagreement on supervisory authority and stare decisis: Justice D’Auria rejects the suggestion that the standards limiting supervisory power or stare decisis should meaningfully impede abolishing the instruction, especially when the court is acting prospectively.
Policy prescription: He would abolish consciousness-of-guilt instructions, or alternatively allow them only when the defense requests them or does not object and the trial court delivers a balanced, neutral instruction reflecting both sides’ characterizations.
Harmless error: He does not have “fair assurance” the erroneous instruction did not substantially affect the verdict; the case was largely a credibility contest, and the split verdict does not reliably show the error was harmless.

III. Analysis

A. Precedents Cited

1. Supervisory authority: prospective rules vs. case-specific reversal

The opinion’s institutional centerpiece is the distinction (and corresponding standards) articulated in State v. Carrion. Justice D’Auria emphasizes Carrion’s two-category taxonomy:

Category 1 (prospective guidance without reversal): The court may “articulate a procedural rule as a matter of policy” without disturbing the conviction. In this category, per State v. Carrion, the court may act when “prudence and good sense so dictate.”
Category 2 (supervisory power as a remedy reversing a conviction): The defendant must show supervisory intervention is “truly necessary” to justify the “extraordinary remedy” of reversal.

Justice D’Auria criticizes the majority for importing the “truly necessary” mindset into the forward-looking question of whether to abolish the instruction. He further relies on the court’s modern efforts to clarify the Carrion distinction, citing State v. Elson and State v. Diaz.

He strengthens the point by canvassing Connecticut’s frequent supervisory use to reform jury instructions—often without granting new trials—including: State v. Medrano (credibility instruction re: defendant’s interest), State v. Ledbetter (eyewitness identification instruction, later “overruled in part” on other grounds by State v. Harris), State v. O’Neil (Chip Smith charge), State v. Aponte (deadly weapon “deemed to have intended” charge), State v. Griffin (two-inference charge), State v. Devalle (reasonable doubt and “ingenuity of counsel”), and State v. Schiappa (reasonable doubt “protect the innocent and not the guilty”).

He also analogizes to supervisory canvass requirements, such as State v. Connor, State v. Gore, and Duperry v. Solnit, and underscores that supervisory power is reserved for serious matters protecting systemic fairness, citing In re Daniel N..

2. Stare decisis and procedural/evidentiary rules

Justice D’Auria frames stare decisis as a tool for stability that must still “serve a justice related end,” quoting State v. Salamon. He argues that stare decisis has not prevented Connecticut from abandoning entrenched instructions and even underlying evidentiary doctrines, highlighting State v. Medrano (ending decades-old credibility language) and State v. Malave (abandoning the missing witness evidentiary rule and instruction in criminal cases).

He adds that stare decisis is weakest for “procedural and evidentiary rules,” invoking Payne v. Tennessee. On this view, eliminating consciousness-of-guilt instructions would be no more jurisprudentially disruptive than the majority’s own reform, which he contends effectively unsettles many prior “no error” consciousness-of-guilt decisions (even if not expressly overruled).

To illustrate the breadth of what the majority’s new restrictions implicitly displace, he lists earlier consciousness-of-guilt cases that approved instructions now disfavored under the majority’s approach, including State v. Cooper, State v. Coward, State v. Luster, State v. Figueroa, and State v. Hines.

3. The “comment on the evidence” critique and out-of-state authority

Justice D’Auria situates consciousness-of-guilt instructions within a broader judicial retreat from “marshalling” evidence. He relies on United States v. Mundy (Second Circuit), which characterizes flight instructions as vestigial and disfavored because judges inevitably exercise judgment and risk invading the jury’s province. He also cites United States v. Robinson (D.C. Circuit) for the proposition that counsel argument is often the best forum for such inferences.

He roots the concern in Connecticut’s own recognition of the trial judge’s influence, quoting State v. Hernandez. He further supports abolition (or strong discouragement) by collecting multiple jurisdictions that discourage or prohibit such instructions: Robbins v. People, Fenelon v. State, Renner v. State, Dill v. State, State v. Marsh, State v. Cathey, State v. Grant, State v. Menard, Hawkins v. State, State v. Jefferson, and Hadden v. State.

4. Harmless error framework and split verdicts

On harm, Justice D’Auria applies the nonconstitutional instructional-error test: whether “it is reasonably probable that the jury was misled,” citing State v. Edwards. He emphasizes that the defendant bears the burden but that the reviewing court must have “fair assurance” the error did not substantially affect the verdict, citing State v. Adam P. and State v. Boyd.

He rejects reliance on a split verdict as a proxy for harmlessness, citing cases rejecting similar inferences: State v. Iban C. (acquittal on more serious charge does not show improper influence was absent) and State v. Maguire. He acknowledges the general presumption that juries follow instructions from State v. Ancona, but argues it does not answer whether the improper instruction misled jurors in evaluating the defendant’s credibility.

B. Legal Reasoning

1. Why the instruction should be abolished (or nearly so)

The opinion advances a policy-based syllogism: consciousness-of-guilt instructions are (i) largely unnecessary because jurors can draw inferences without judicial prompting, (ii) duplicative of general instructions on circumstantial evidence and credibility, and (iii) uniquely risky because they spotlight an inference that goes to the ultimate issue—guilt—under the imprimatur of the court.

Justice D’Auria treats the instruction as a judicial endorsement problem. Even if couched in “may but need not” language, it invites jurors to treat the defendant’s conduct as especially probative because the judge has singled it out. He analogizes to State v. Malave’s warning about the difference between what jurors infer on their own and what they infer when the judge highlights the inference.

He also criticizes the administrability of the majority’s retained-in-theory/rare-in-practice model. The majority requires the state (when the defense objects) to justify why the jury needs special guidance beyond general instructions; Justice D’Auria argues neither the state nor the majority can provide concrete examples, and the state conceded at argument that the instruction is not “critical.” On his account, this predicts continued litigation: appeals will turn on whether a trial judge correctly applied an amorphous necessity balancing test.

2. Supervisory authority and stare decisis as non-barriers

Justice D’Auria’s institutional claim is narrow but significant: if the court can use supervisory authority to prospectively “clarify,” “modify,” or “restrict” an instruction, it can also abolish it—because both are prospective judicial-policy choices judged by “prudence and good sense” under State v. Carrion.

Likewise, stare decisis cannot logically bar abolition while permitting near-elimination via a new rubric. The majority’s approach, he argues, already departs from prior permissive practice; thus, invoking stare decisis to retain “fragments” of the instruction is internally inconsistent.

3. Why the error was harmful here

On the record, Justice D’Auria stresses that the instruction was not merely unnecessary; it was skewed. The trial court told jurors the defendant “change[d] his story” about whether he went to the complainant’s home, based on workplace interview evidence that the majority itself deemed “far less definitive.” That “pejorative connotation” risked signaling a judicial endorsement of the state’s credibility attack.

He rejects three common harmlessness arguments:

General instructions cured it: He argues the opposite—adding a specific consciousness-of-guilt instruction on top of general credibility/circumstantial-evidence instructions heightens emphasis rather than neutralizing it.
The instruction’s own “limiting caveat” cured it: He argues the model language does not fix the problem of judicially adopting the state’s characterization (“changed his story”).
Split verdict shows careful deliberation: He argues the acquittals do not prove the jury was not improperly influenced on the counts of conviction; it may simply show the jury did not fully credit the complainant, and the improper instruction could have been the “tipping” factor on the lesser counts.

Because the defendant’s account plausibly explained some physical evidence (contact, DNA, and injuries from being pushed away), and because the jury acquitted on the most serious charges, Justice D’Auria cannot reach “fair assurance” that the instruction did not substantially affect the guilty verdicts.

C. Impact

Although a concurrence/dissent does not itself establish binding doctrine, this opinion is positioned to influence Connecticut practice and appellate argument in three ways:

Supervisory authority framing: It offers litigants a clean template (from State v. Carrion) for arguing that prospective abolition of a problematic instruction requires only “prudence and good sense,” not the heightened “truly necessary” justification tied to reversal.
Stare decisis limits in instruction reform: By anchoring jury-instruction reform in State v. Medrano and State v. Malave, it supports future claims that entrenched instructions can and should be eliminated when they no longer serve fairness.
Harmless-error skepticism in credibility cases: It warns that when a case turns on witness credibility, judicial emphasis on a “guilty conscience” inference—especially using pejorative phrasing—can be outcome-affecting, and a split verdict does not reliably demonstrate harmlessness (State v. Iban C.; State v. Maguire).

Practically, the opinion also anticipates an appellate “feedback loop”: as long as the instruction survives in any form, appellate courts will repeatedly litigate (a) whether evidence justifies the instruction, (b) whether the wording was neutral, and (c) whether any error was harmless—until abolition becomes the cleaner systemic solution.

IV. Complex Concepts Simplified

Consciousness of guilt instruction: A jury instruction telling jurors they may infer guilt from certain post-event conduct (e.g., false statements, flight, concealment). The critique here is that it “spotlights” one inference and can make jurors overweight it because it comes from the judge.
Supervisory authority: The Connecticut Supreme Court’s power to regulate procedures in trial courts to protect fairness and the integrity of the judicial system. Under State v. Carrion, it is easier to use this power prospectively (setting rules for the future) than to reverse a conviction as a remedy.
Stare decisis: The principle of adhering to precedent for stability. The opinion emphasizes it is not absolute—especially for procedural/evidentiary rules—and must serve justice (State v. Salamon; Payne v. Tennessee).
Harmless error (nonconstitutional): Even if the trial court erred, a conviction stands unless it is “reasonably probable” the jury was misled (State v. Edwards). Appellate courts look for “fair assurance” the error did not substantially affect the verdict (State v. Boyd).
Split verdict: A jury convicts on some counts and acquits on others. The opinion warns that this does not automatically show an error was harmless; an error may still have influenced the counts of conviction (State v. Iban C.).

V. Conclusion

Justice D’Auria’s concurrence/dissent in State v. Kim is both a doctrinal clarification and a policy critique. Doctrinally, it underscores that prospective supervisory reforms are governed by “prudence and good sense” (per State v. Carrion) and that stare decisis is not a meaningful barrier to instruction abolition in the procedural/evidentiary sphere. As a matter of policy, it contends consciousness-of-guilt instructions are unnecessary, duplicative, and unfairly prejudicial because they judicially elevate an inference closely tied to guilt. Finally, applying Connecticut’s harmless-error standards, it argues the specific, unbalanced instruction here—pejoratively asserting the defendant “changed his story”—likely affected a credibility-driven verdict and thus warrants a new trial.