Fake Cop “Experts”: Leading a Double Life

June 6, 2017
By Hanging Out with Carl Gunn


  • The Ninth Circuit has recognized that “dual role” testimony, where a cop testifies as both a lay witness and an expert witness raises special problems that a trial court has to guard against.
  • While such testimony is not categorically prohibited, the court does have to take protective measures, such as giving a cautionary instruction, clearly separating the lay testimony from the expert testimony, and/or assuring there’s an adequate foundation for the expert foundations.
  • Two cases have found plain error in the admission of such testimony, so it’s an issue to watch for in your cases.



I talked in last week’s post about some bad cases and a few good cases considering certain types of cop “expert” testimony.  There’s another line of cases – which are at least potentially good – that considers a context problem that arises in some cop “expert” cases.  That’s where the cop testifies as both a lay witness about some facts and an expert witness about others.  The Ninth Circuit has recognized this raises special problems which a district court has to be careful to guard against.  The court first recognized this in United States v. Freeman, 498 F.3d 893 (9th Cir. 2007), where it listed the problems as follows:

First, by qualifying as an ‘expert,’ the witness attains unmerited credibility when testifying about factual matters from first-hand knowledge.  Second, it is possible that expert testimony by a fact witness or case agent can inhibit cross-examination . . . [because a] failed effort to impeach the witness as expert may effectively enhance his credibility as a fact witness.  Third, when the prosecution uses a case agent as an expert, there is an increased danger that the expert testimony will stray from applying reliable methodology and convey to the jury the witness’s sweeping conclusions about appellants’ activities, deviating from the strictures of Rules 403 and 702.  Fourth, a case agent testifying as an expert may lead to juror confusion because [s]ome jurors will find it difficult to discern whether the witness is relying properly on his general experience and reliable methodology, or improperly on what he has learned of the case.  Finally, when a case agent/expert strays from the scope of his expertise, he may impermissibly rely upon and convey hearsay evidence.  In doing so, the witness may also run afoul of the Sixth Amendment Confrontation Clause.

Id. at 903 (internal quotation marks and citations omitted).  See also United States v. Vera, 770 F.3d 1232, 1242 (9th Cir. 2014) (citing and summarizing concerns identified in Freeman).

The court nonetheless concluded that such testimony “is not so inherently suspect that it should be categorically prohibited.”  Freeman, 498 F.3d at 904.  On the other hand, it emphasized “the necessity of making clear to the jury what the attendant circumstances are in allowing [the] agent to testify as an expert.”  Id.  And later cases have pointed to at least three things a trial court allowing such testimony should consider.  First, the court should make the distinction between expert and percipient testimony clear by separating the expert testimony from the non-expert testimony.  See, e.g., United States v. Anchrum, 590 F.3d 795, 803 (9th Cir. 2009).  Second, the court should specifically instruct the jury at the time of the testimony on the difference between percipient and expert testimony.  See, e.g., United States v. Martinez, 657 F.3d 811, 817 (9th Cir. 2011).  Compare Vera, 770 F.3d at 1246 (noting absence of instruction that agent’s lay opinion testimony was not based on scientific, technical, or other specialized knowledge).  See also Manual of Model Criminal Jury Instructions for the District Courts of the Ninth Circuit § 4.14A (2010 ed. 12/2016 update) (model instruction).  Finally, the court should require the government to lay an adequate foundation for the expert opinions the agent gives.  See Vera, 770 F.3d at 1247 (noting as additional reason for finding plain error “the general failure to assure an adequate foundation for [the agent’s] opinions”).

At least two Ninth Circuit cases have found plain error where “dual role” agent testimony was allowed without these protections.  In Vera, the court found plain error because the district court had both failed to instruct the jury on how to evaluate the agent’s testimony and failed to require a sufficient foundation for his opinions, see id., 770 F.3d at 1243, and found it prejudicial and requiring reversal of the jury’s finding of drug quantity, though not the underlying conspiracy conviction, see id. at 249.  In United States v. Torralba-Mendia, 784 F.3d 652 (9th Cir. 2015), the court also found plain error in the failure to instruct the jury on how to evaluate such testimony, see id. at 659, though it found the error non-prejudicial because the government had bifurcated the agent’s testimony and the agent had provided an adequate foundation for his opinions, see id. at 661-62.  This latter case suggests there may need to be more than one deficiency in the district court’s handling of such testimony, and the first case, Vera, suggests – as makes sense in plain error analysis, or even harmless error analysis – that the importance of the evidence and closeness of the case matter.

In any event, keep an eye out for this issue.  First, argue that your court shouldn’t allow such dual role testimony because it needlessly creates problems, pointing to the list of potential problems identified in Freeman.  Second, if the court allows the testimony anyway, remind it of “the necessity of making clear to the jury what the attendant circumstances are in allowing [the] agent to testify as an expert,” and ask it to (1) give an instruction at the time of the testimony, (2) make the government clearly separate the expert and non-expert testimony, and (3) assure there’s an adequate foundation for the opinions the expert gives.  Expressly asking for these things (1) may limit the testimony to some extent and (2) will keep a future appellate colleague from having to argue the more difficult plain error standard of review.