Labor & Employment Law Daily Wrap Up, NAA CORNER—Assessing credibility in ‘he said/she said’ cases, (Aug 19, 2025)

By Arbitrator Gina Roccanova and Arbitrator Robert Hirsch, NAA Member

Cases that turn on assessments of two witnesses’ competing accounts can be particularly difficult.

From time to time, arbitrators, investigators, and other finders of fact must decide cases wholly or mostly on the basis of witness testimony. Cases that turn on assessments of two witnesses’ competing accounts can be particularly difficult. In those matters it's not sufficient to say, "well, it's he said/she said, therefore the allegations can't be proven." Those cases require an assessment of the credibility of each witness.

The following hypothetical, based loosely on actual cases, illustrates some of the challenges in conducting credibility assessments:

Grievant worked for a company that provided services to clients in their homes. Ms. X, one of Grievant's clients, reported to Grievant's employer that Grievant had sexually assaulted her during a home visit. No one else was present during the alleged assault. After an investigation the employer fired Grievant.

At the grievance arbitration hearing, which was held via videoconference, Ms. X gave detailed testimony about the alleged assault, and her account was consistent with her original report 10 months earlier. Although she seemed nervous and uncomfortable with the hearing process itself, she unflinchingly confirmed several unflattering facts, including a history of drug abuse and a criminal record that included convictions for prostitution, drug-related charges, and robbery. Ms. X even volunteered details that did not show her in the best light.

Grievant was a family man and a long-term employee who seemed friendly, relaxed, and confident on the witness stand. His recollection of the date in question was less detailed than that of Ms. X, but he stated that he had no reason to recall more details of that day because nothing untoward had occurred.

The Employer introduced several pieces of circumstantial evidence, including (1) Grievant’s own records showing an inordinate amount of time between the beginning of the visit with Ms. X and the next visit recorded, a gap for which Grievant had no explanation; (2) testimony from a neighbor of Ms. X who observed her visibly upset and crying after the visit.

Grievant’s lawyer attacked Ms. X’s credibility by highlighting her criminal background and addiction issues.

This scenario gives rise to several questions:

1. To what degree does witness demeanor - particularly body language - matter?

In the Spring of 2020, when the COVID-19 pandemic forced many arbitration proceedings to take place via videoconference, arbitrators and advocates raised the concern that videoconference hearings would not permit an adequate assessment of credibility because the finder of fact could not observe the witnesses fully.[1] Those discussions brought to the fore the extent to which arbitrators rely on demeanor, particularly physical cues, when assessing a witness’ truthfulness. When finders of fact observe witnesses with nervous hands, darting or averted eyes, tapping feet, crossed arms, etc., it’s easy to attribute those mannerisms to dishonesty or a desire to avoid questions. But social scientists and legal scholars have long cautioned against reliance on “external appearance and physical cues” in assessing credibility.[2] Those nonverbal cues - which are aspects of but not the entirety of the definition of “demeanor” - may just as likely result from anxiety about being in an unfamiliar situation, or ADHD, or the culture that witness was raised in. Moreover, focusing on body language can draw focus away from the aspects of demeanor, specifically vocal cues like tone of voice and hesitation, that more accurately track with truthfulness.

2. To what degree should the witness’ background be considered?

Evidence of a witness’ prior felony conviction is generally admissible under the Federal Rules of Evidence. In the scenario above, the Grievant’s counsel leaned in to a widely held perception that people who have committed serious violations are less likely to respect the legal system and the witness’ oath specifically. However reliable that assumption may be overall - and that issue is the subject of debate in both legal and social science circles[3] - it is particularly problematic in this context. Gender, race, sexual orientation, addiction, poverty, and prior experience with the legal system impact individuals’ willingness to report crimes against them, as well as whether they are believed when they do report. This in turn may make them more vulnerable to abusers.[4] In this context, the same factors that the Grievant’s counsel zeroed in on as reflecting negatively on Ms. X’s credibility (i.e., her status as a former sex worker and former addict with a criminal history and other negative experiences with the legal system) made it more likely that she would be a target of exactly the type of conduct that she complained about.

3. What should finders of fact consider in assessing credibility?

How, then, should a finder of fact assess the relative credibility of competing testimony? In its guide for prevention of workplace and housing harassment, the California Civil Rights Department has provided a helpful and well researched list of factors to be used in assessing credibility.[5] Although created for the specific context of harassment investigations, these factors can be applied more broadly in assessing credibility. They include:

1. Inherent plausibility. Does the story make sense?

2. Motive to lie. Is the witness biased, or do they have some stake in the outcome?

3. Corroboration. Even if the central allegations cannot be corroborated, do other witnesses or documents corroborate other aspects of the witness’ testimony, such as timelines?

4. Ability to perceive, recall, or communicate. For example, was the witness in a physical location where they could observe the incident they are testifying about? Is the witness’ memory spotty?

5. History of honesty/dishonesty. Has the witness been dishonest in similar situations? Has the witness been forthcoming about unflattering facts?

6. Habit/consistency. Is the testimony about an individual consistent with the way they act habitually, e.g., this person always takes off his shoes as soon as he enters a home?

7. Inconsistent statements. Is the witness’ account internally consistent and consistent with previous statements about the same incident? If not, can the inconsistency be explained?

8. Manner of testimony. This is the aspect of demeanor evidence that is reliable, such as whether the witness answers questions without hesitation or evades direct responses.

Although the guidelines also list “demeanor” as a factor, they caution against over-reliance on that factor given its unreliability.

Conclusion

In the case described above, the arbitrator upheld the termination, determining that Ms. X was a woman who had had enough abuse in her life and wanted to stop an individual who posed a danger to others. Rather than her demeanor, her testimony - its specificity - and her apparent candor in the face of the embarrassing details of her life - made her believable. The circumstantial evidence, which strongly supported her version of events, was important as well.

This scenario illustrates both the difficulty inherent in resolving competing versions of events and the pitfalls of relying on credibility factors that have long been applied without questioning their reliability. A careful and thorough application of the other, more reliable factors is far more likely to lead to a just result.

About the authors. Gina Roccanova is a professional neutral arbitrator, mediator, and hearing officer concentrating in labor and employment disputes. She has significant experience with public agencies, public safety, higher education, K-12, health care, transit, nonprofits, cultural institutions, performing arts, bioscience, utilities, and media. Robert Hirsh has over 40 years of participation in arbitration and mediation, as a neutral and an advocate, in labor/employment and business/commercial matters. Industries include real estate, building and construction, hospitality, education, health care, automotive, trucking/transportation, furniture, meat packing, bakery, office clericals, horse racing, pension-fiduciary, and investments.

Footnotes

1 On the other hand, many arbitrators believe that video hearings can provide an even greater opportunity to observe witness’ demeanor.

2 See,. e.g., Jeremy A. Blumenthal, A Wipe of the Hands, A Lick of the Lips: The Validity of Demeanor Evidence in Assessing Witness Credibility, 72 NEB. L. REV. 1157 (1993) and research cited therein.

3 For a broad critique of the Rules of Evidence use of status as a poor proxy for truthfulness, see Julia Simon-Kerr, Credibility by Proxy, 85 GEO. WASH. L. Rev. 152 (2017).

4Deborah Tuerkheimer, Incredible Women: Sexual Violence and the Credibility Discount, 166 U. PA. L. REV. 1 (2017-2018).

5 DFEH-Workplace-Harassment-Guide.pdf

Companies: National Association of Arbitrators

MainStory: Arbitration SexualHarassment EvidenceDiscovery