Guilt by Association As Circumstantial Evidence

When Does Circumstantial Evidence Go Too Far?

A guilt by association with others is a simple but dangerous shortcut.

The prosecution likes to argue that:

1. Witness A is a friend or associate of the accused who committed misconduct;
2. therefore, the accused probably committed it too.

The shortcut can be especially difficult to recognize because it often appears dressed up as circumstantial evidence. That’s what prosecutor’s call it.

Military appellate courts have rejected guilt by association for a long time. Circumstantial evidence is okay. But that evidence still has to show it was the accused who did the actual crime. It is not a rule of evidence, but an argument based on circumstances.

The rule is old

The Court of Military Appeals confronted the problem more than seventy years ago. In United States v. Jacobs, 1 C.M.A. 209, 2 C.M.R. 115 (1952), the court warned against allowing criminal responsibility to rest on association with others. The court described guilt by association as a principle “alien to American standards of justice.” The court reiterated the point in United States v. Adkins, 5 C.M.A. 492, 18 C.M.R. 116 (1955).

The principle has not disappeared. In United States v. Diffoot, 64 M.J. 116 (C.A.A.F. 2000), the Court of Appeals for the Armed Forces considered a prosecutor’s argument that essentially invited the members to infer the accused’s guilt from his association with known criminals. CAAF found the argument prejudicial.

The underlying proposition is straightforward: the Government must prove what the accused did, not what the people around the accused did.

That does not mean association is irrelevant. It can be powerful evidence when it logically tends to establish knowledge, intent, motive, opportunity, or participation in a particular offense. The problem arises when association itself supplies the missing proof.

The difference between evidence and inference

This distinction matters because criminal cases routinely depend on circumstantial evidence.

Suppose three people are involved in an armed robbery. Evidence that the accused was with the other two immediately before and after the robbery may be highly relevant. Evidence that the accused discussed the robbery, supplied the vehicle, divided the proceeds, or fled with the others may make the inference of participation even stronger.

But consider a different proposition:

Two people associated with the accused committed the robbery. The accused associated with them. Therefore, the accused probably participated in the robbery.

That is not really evidence of the accused’s participation. It is an inference based principally on association.

The distinction can be subtle. A factfinder is entitled to draw reasonable inferences from circumstantial evidence. But the inference must have a logical connection to the fact the Government must prove.

This is why “common sense” is not an answer to every evidentiary question. Common sense permits reasonable inferences. It does not eliminate the Government’s burden of proof.

How the problem comes up at court-martial

Guilt by association may arise in several forms.

Conspiracy and concerted action. Association with members of a conspiracy can be relevant, but association alone does not establish an agreement or participation in it.

Drug offenses. Being present with people who possess or use drugs does not, by itself, establish that the accused possessed the drugs.

Sex offenses. A relationship with another person who engaged in sexual misconduct cannot substitute for proof of the accused’s own conduct.

Weapons offenses. The fact that other people in the accused’s group carried weapons, for example, does not necessarily establish that the accused carried one.

Other misconduct. The same problem can arise whenever the Government relies on another person’s conduct as a stepping stone toward proving the accused’s conduct.

The common feature is the same: the Government has evidence of someone else’s conduct and needs an additional inferential step to connect that conduct to the accused.

Sometimes that step is justified. Sometimes it is not.

Vongphachanh: association as circumstantial evidence

A recent case illustrates the problem.

In United States v. Vongphachanh, C.A.A.F. Dkt. No. 26-0292/AF, the accused was convicted of, among other offenses, unlawfully carrying a concealed weapon. There was no dispute that he possessed and fired a handgun. The more difficult question was whether the Government proved that he carried it concealed and unlawfully.

The Air Force Court of Criminal Appeals considered the evidence in conducting its Article 66(d) factual-sufficiency review. Among other things, the court noted that two of the accused’s friends were carrying concealed weapons because they anticipated a confrontation. The court concluded that “it stands to reason” that the accused likewise kept his weapon concealed.

That reasoning presents an interesting question.

The fact that the accused’s companions concealed their weapons may be relevant. It provides some context for the circumstances in which the accused possessed his own weapon. But it does not establish that he concealed his weapon.

The inferential chain is:

1. Two people associated with the accused carried concealed weapons.
2. They anticipated a confrontation.
3. The accused was with them.
4. Therefore, the accused probably also concealed his weapon.

The first three propositions may be supported by evidence. The fourth is an inference. The question is whether the evidence adequately supports that inference or whether association has effectively supplied the missing proof.

That distinction becomes more important when the direct evidence is weak. In Vongphachanh, the defense challenged the characterization of testimony from a principal witness who said he had not seen a weapon on the accused. The defense argued that the witness’s inability to see a weapon did not establish that the weapon was concealed.

That is a different proposition from saying that the weapon was concealed because the accused’s companions concealed theirs.

The question on appeal

The point is not that a conviction can never rest on circumstantial evidence. It plainly can.

Nor is the question whether an appellate court would have reached the same conclusion as the factfinder. Under Article 66(d), the Court of Criminal Appeals performs its own factual-sufficiency review within the statutory framework.

The more fundamental question is:

Did the evidence actually establish the element of the offense that the prosecution must prove

That question matters particularly on factual sufficiency review. An appellate court should not merely identify facts consistent with guilt. It must consider whether the Government proved the required elements and whether the evidence, taken together, supports the finding of guilt.

If the decisive inference is that other people did it, and the accused was with them, the appellate court should take a closer look.

What should the defense counsel consider

Guilt by association is often easier to recognize in argument than in the evidence itself.

Defense counsel should listen for formulations such as:

• “They were all together.”
• “Everyone else had one.”
• “His friends were doing it.”
• “He ran with that crowd.”
• “They were all involved.”
• “It would make sense that he did the same thing.”
• “Why else would he be with them?”

None of those statements necessarily constitutes improper argument. The question is what proposition the Government asks the factfinder to infer from them.

Counsel should ask a more precise question:

What evidence proves that the accused himself did it

If the answer is independent evidence of the accused’s conduct, association may simply be part of a larger circumstantial case.

If the answer ultimately circles back to the conduct of the accused’s companions, the defense may have identified a genuine evidentiary weakness.

The larger point

Guilt by association is not a rule against considering relationships. It is a rule against confusing association with culpability.

The Government is entitled to prove a case circumstantially. Factfinders are entitled to draw reasonable inferences. But the inference must bridge the evidentiary gap between the accused and the offense.

That is particularly important in military justice, where the Government frequently presents cases built from surrounding circumstances rather than eyewitness testimony.

The right question is not whether the accused was associated with people who committed misconduct.

It is whether the evidence establishes what the accused did.

That distinction can be the difference between legitimate circumstantial proof and guilt by association.

More from Cave & Freeburg on trials and courts-martial appeals.

1. “Other acts—predisposition—Rule 404(b)” is useful in discussing the basic prohibition against using other acts to establish propensity and explains how defense counsel can challenge such evidence. Philip D. Cave. Other acts—predisposition—Rule 404(b)

2. “404(b) evidence of common scheme / plan” The post also discusses the M.R.E. 403 balancing requirement. Philip D. Cave. 404(b) Evidence of Common Scheme / Plan.

3. “Menard and the Misuse of M.R.E. 404(b)(2) as Evidence of a Common Plan or Scheme” discusses the danger of transforming character or general behavioral patterns into a supposedly legitimate nonpropensity inference. Philip D. Cave. Menard and the Misuse of M.R.E. 404(b)(2)

4. “Modus Operandi evidence” is useful for the distinction between legitimate use of similarities and impermissible propensity reasoning. It specifically discusses M.R.E. 404(b) and the Reynolds framework. Philip D. Cave. Modus Operandi Evidence