Four recent decisions by federal courts show just how much weight an employer’s investigation can hold in litigation, for better or worse. Investigation quality does not operate on a simple pass/fail switch. It moves along a spectrum, and where an employer’s investigation lands on that spectrum — from “feeble” to “flawed” — can decide a case.

A Feeble Investigation Can Cost More Than the Underlying Claim

In EEOC v. SkyWest Airlines, Inc., the 5th Circuit affirmed a jury verdict awarding compensatory and punitive damages to a parts clerk who was subjected to extreme sexual harassment by her coworkers at DFW International Airport. The employer investigated after the employee complained to HR, but the court found the investigation itself undermined the company’s defense to punitive damages. Normally, the investigator would interview every witness, but in this case interviewees were randomly selected. Also, one witness admitted to ongoing sex talk in the workplace, but the investigator asked no follow-up questions about what that meant. The investigator did not try to obtain available corroborating evidence, such as photographs connected to one incident. At the end of the investigation, the investigator issued written warnings to several employees but no discipline at all to a maintenance supervisor directly involved in the harassment. The investigator also did not contact the employee about the conclusion of the investigation.

The court held these deficits in the investigation were enough for the jury to conclude the investigation was, in the court’s words, a feeble attempt to uncover the truth, which defeated the employer’s good-faith defense to the employee’s claim for punitive damages.

Missing Files and Skipped Witnesses Can Send a Case to a Jury

In Brenyah v. Columbia Hospital Corp. of Bay Area, the 5th Circuit reversed summary judgment on a nurse’s Title VII and Section 1981 hostile work environment claims after identifying several problems with the employer’s investigation into her complaints of race and national origin harassment. The employer argued that it had investigated promptly by interviewing employees on the unit and coaching nurses who made the comments at issue.

However, the employee pointed to a different picture that included discrepancies between the two investigators’ testimony and their interview records, a claim by one investigator that an investigation file and interview notes existed even though none were produced in discovery, a black employee on the unit who was never interviewed despite the investigator’s assertion that he had been, corroborating statements that were left out of the investigation summary, and a departure from the employer’s usual practices of creating interview summaries and allowing employees to review and sign their own interview statements.

The court held that the employee had pointed to enough evidence to show that the investigation was not sufficiently prompt or thorough, and the employer’s actions were not calculated to end the harassment. Because the record left a genuine dispute about whether the employer met that standard, summary judgment on the hostile work environment claims was reversed.

Model Investigation: An Independent Board Investigates for Two Months

In Sotonwa v. Collins, the 5th Circuit affirmed summary judgment for a Veterans Affairs medical center after it demoted a physician following sexual harassment complaints from two employees after their employee physicals. The VA convened a four-member investigative board with members who had no personal relationship to the physician. Over two months, the board conducted interviews of more than 20 employees. The board found the physician’s exams were inappropriate but did not rise to the level of harassment, though the interviews independently documented a pattern of abrasive, bullying leadership and led to a recommendation of sensitivity and other training.

A new supervisor then independently reviewed the board’s report, personally observed the physician’s conduct, and wrote a detailed memo recommending demotion. The ultimate decision maker reviewed the full investigative file along with the physician’s own written rebuttal before making the final call to demote. When the physician made a claim of race, color, and national origin discrimination, the VA moved for summary judgment, which the trial court granted. The 5th Circuit held that a thorough, independent, and well documented investigation, followed by a layered chain of independent review, was strong evidence of a legitimate, nondiscriminatory reason for an adverse action, and the physician’s subjective belief that the process was tainted could not overcome that evidence.

Good-Faith Reliance: The Investigation Does Not Have to Be Perfect

In Brauckmiller v. University of Texas San Antonio, a federal district court granted summary judgment for a university that terminated a senior manager after its Equal Opportunity Services office investigated a subordinate’s complaint that the manager sent her unwanted messages and comments on her personal cell phone and social media accounts. The investigators met with both the manager and the complaining employee, interviewed eight additional witnesses, and reviewed the submitted and independently gathered evidence. They found the complainant’s allegations credible because they were corroborated by witnesses and objective evidence, and they concluded the manager had violated the university’s “Other Inappropriate Sexual Conduct” policy.

The manager argued the investigation had been biased, that evidence had been altered or concealed, and that there was a conspiracy among the HR employees to villainize him. The court rejected all of it, explaining that the question on summary judgment was not whether the investigation’s findings were accurate but whether the university was entitled to rely on the investigation and, if not, whether its bad-faith reliance on the investigation was pretext for discrimination or retaliation.

The court held that when a decision-maker relies on the results of an investigation, evidence that the investigation was flawed or reached inaccurate conclusions does not by itself show the decision-maker’s reliance was pretext for discrimination, and there was no evidence that UTSA relied on the report in bad faith as a pretext for discrimination. As the court put it, even if the investigation report contains some inaccuracies, an employer may fire an employee for a good reason, a bad reason, a reason based on erroneous facts, or for no reason at all, so long as the real reason is not a discriminatory one.

Takeaways for Employers

• Interview everyone with relevant knowledge, not a convenient subset. Randomly selecting witnesses or skipping a witness who could corroborate a complaint does not look good to a court or a jury.
• Follow up on ambiguous or incomplete answers, and corroborate allegations with witnesses and objective evidence wherever possible. A witness admission that gets no follow-up question, or evidence that is never requested, can later be read as an investigator who did not want to know more.
• Document every step contemporaneously, and preserve what is documented. Claiming that a file and notes exist, without producing them in discovery, may be treated as meaningful evidence against the employer.
• Apply established procedures consistently. A departure from an employer’s own normal investigative practice can itself become evidence against the organization.
• Use independent investigators with no personal stake in the outcome, and build a layered, independent decision chain on top of the investigation. That layered review can make the ultimate decision far more defensible.
• Match the disciplinary response to the seriousness of the conduct. An investigation that results in no discipline for the individual most directly involved in the alleged misconduct, while others receive lesser consequences, invites scrutiny of whether the response was truly calculated to end the problem.
• Tell the complainant how the investigation came out. This is required in California, but should be a part of every investigation. The person who made the complaint should know whether it was substantiated and whether remedial action was taken. 

Investigation quality is not a background administrative task. It is evidence, and courts are treating it that way. Organizations that build independent, consistent, well-documented, and thorough investigation practices, with layered review before the final decision, are far better positioned to defend their decisions.