Subordinate Bias Liability: The 'Cat's Paw' Doctrine

The courts thus have endorsed the notion that an employer should not be able to avoid liability through willful blindness, or acting as the proverbial ostrich with its head in the sand. It is also clear that a biased employee with no final decision making power can affect an ultimate employment decision. Manufacturing performance issues, selective reporting, and other actions by a biased subordinate can all lead to an adverse employment action by an impartial decision maker based on tainted facts. However, though the 'cat's paw' doctrine is generally accepted, its scope and, specifically, the question of what degree of control a subordinate must have over the decision maker and ultimately over the adverse action, greatly differ from court to court.

There are three standards that have been used in answering those questions. The Fifth Circuit applies the most liberal test by considering whether the subordinate 'possessed leverage, or exerted influence, over the titular decision maker.' Russell v. McKinney Hosp. Venture, 235 F.3d 219, 227 (5th Cir. 2000). The Fifth Circuit and Seventh Circuit have held that 'summary judgment generally is improper where the plaintiff can show that an employee with discriminatory animus provided factual information or other input that may have affected the adverse employment action.' Dey v. Colt Constr. & Dev. Co., 28 F.3d 1446, 1459 (7th Cir. 1994) (citing Shager, 913 F.2d at 405). Thus, providing biased 'factual information or other input' is sufficient to attach liability so long as it 'may have affected' the outcome.

The Fourth Circuit in Hill, 354 F.3d at 289, articulated a standard at the other extreme. That case held that though the subordinate does not need to be the 'formal decision maker,' he or she has to be, 'the one 'principally responsible' for, or the 'actual decision maker' behind, the action.' The Fourth Circuit declined to extend liability upon proof that the subordinate 'had a substantial influence on the ultimate decision or because he has played a role, even a significant one, in the adverse employment decision.' Id. at 291. This strict interpretation would allow an employer to evade liability with willful blindness because an employee could suffer an adverse action based solely on discriminatory bias or animus of a non-decision maker; no liability would attach because the biased individual did not have complete control over the formal decision maker. That interpretation is also the least popular; the dissent in Hill noted that the majority's decision 'puts us at odds with virtually every other circuit.' 354 F.3d at 299.

The Seventh and Tenth Circuits have required a causal link between the discriminatory animus of the non-decision maker and the adverse action. In their formulation, if a decision maker is dependent upon a biased employee for information, a mere 'paper review' of that information will not be enough to shield a company from liability. However, if the decision makers perform more than a perfunctory review and conduct their own independent research, there would be no finding of discrimination. Hence, the company can be excused even if a good portion of the information justifying the decision came from the biased employee.

A case involving that standard was scheduled to be argued before the Supreme Court last month, but was withdrawn at the last moment. Though the Tenth Circuit case will not be heard by the Court, an analysis of the facts of the case is useful in illustrating the 'cat's paw' issue. Furthermore, that case provides examples of how a firm can structure its human resources and management processes to prevent violations of Title VII.

In EEOC v. BCI Coca-Cola Bottling Co., of Los Angeles, 450 F.3d 476 (10th Cir. 2006), Peters, an African-American employee, was dismissed from his job as a merchandiser at BCI's facility in Albuquerque, NM. He reported to a Hispanic district sales manager, Grado. Peters' day-to-day supervisor was Katt, who also reported to Grado, the allegedly biased subordinate. The EEOC cited several incidents that purportedly demonstrate that Grado treated African-American employees more harshly than white and Hispanic employees.

The incident at issue involved a weekend that Peters was not scheduled to work, but due to an employee shortage, was told to come in. Grado found out that Peters was not planning to work and called a human resources supervisor, Edgar, based in Phoenix, more than 450 miles away. Edgar advised Grado to tell Peters that failure to comply with the order to report would be insubordination that could lead to dismissal. Only Edgar had the authority to dismiss an employee.

That Friday, following Edgar's instructions, Peters and Grado had a discussion about his working over the weekend and Peters' responded with, '[D]o what [you've] got to do, and I'll do what [I've] got to do.' Peters was not feeling well and went to a medical clinic on Saturday and called in sick to Katt. Katt approved his absence and then tried, but was unsuccessful, in advising Grado.

On Sunday, Pederson, the on-site human resources representative, told Edgar that Peters had a two-day suspension for an infraction two years earlier. Edgar was not told that Peters did not show up at work on that earlier occasion because he was a pallbearer in his fianc'e's son's funeral and Peters was not given permission to attend because the deceased was not his biological son. On Monday, Edgar told Pederson and Grado to terminate Peters' employment. Edgar asserts that the firing decision was based on the insubordination that Peters allegedly exhibited during his Friday discussion with Grado. It is undisputed that at the time of the termination decision neither Edgar nor Pederson knew Peters' race.

The district court dismissed the case on summary judgment because the decision maker conducted 'independent research,' that is, that there was an independent review of the personnel file to find out about the previous infraction. The Tenth Circuit reversed, finding this research insufficient and held that there was a factual issue as to whether there was a causal link between Grado's allegedly biased actions and the termination, thus warranting a trial.

Given the disagreement as to the applicable standard, the Supreme Court will likely consider a similar case in the future. In the meantime, as the BCI case teaches, companies should establish a mechanism to investigate dismissal recommendations and use independent research that is more than perfunctory. Relying on information from only one employee can be a bad idea. In BCI, the human resources manager, the decision maker, did not talk to the aggrieved employee to get his side of the story. Nor did the human resources decision maker talk to the employee's immediate supervisor to obtain more information. All the information she used in making the final decision came from the allegedly biased manager. The independent examination extended only to Peters' personnel file, not from an inquiry into the subsequent events. Arguably, the Tenth Circuit might have dismissed the claim if either of the human resource personnel had talked to the aggrieved employee or even to his immediate supervisor.

Employers have asserted that this investigative standard would carry a significant burden. Good practice, however, dictates that decision makers, to avoid the possibility of bias, should get the employee's side of a disputed situation and, to the extent possible, get input from other supervisors and managers. Moreover, the additional burden to investigate by speaking to several individuals may not be great, especially where the proposed action is termination, the most serious sanction an employer can take. It is unclear in BCI how burdensome a more thorough process would have been. Would it have been significantly more time consuming for either of the human resource personnel to speak to the employee or his other supervisor directly?

Though BCI was withdrawn, there is still the open question to which the Supreme Court granted certiorari: 'Under what circumstances is an employer liable under federal anti-discrimination laws based on a subordinate's discriminatory animus when the person who actually made the adverse employment decision admittedly harbored no discriminatory motive toward the impacted employee?' Given the disparate views of the circuits, this is an area that is ripe for clarification and one that the Supreme Court will likely address in the near future. In the meantime, by engaging in independent research, a company can help protect employees' rights while minimizing its exposure to liability stemming from a single supervisor's biased recommendation.

Debra L. Raskin is a partner at Vladeck, Waldman, Elias, & Englehard, P.C., and an adjunct faculty member at Columbia Law School. She has lectured and written on employment law matters for the New York State Bar Association and the American Law Institute/American Bar Association, among other groups. She can be reached at [email protected]. Jamie Ostrow worked as an associate in the London office of Freshfields, Bruckhaus Deringer and is currently a consulting attorney with Vladeck in New York.

Everyone knows that a manager who expresses discriminatory views and then fires or disciplines an employee belonging to the disfavored group may create a claim against the employer. But what happens when an unbiased manager relies on the recommendation of another supervisor who, unbeknownst to the decision maker, is a raging bigot?

Many courts have recognized claims under Title VII for subordinate bias based on 'cat's paw' or 'rubber stamp' theories of liability. The 'cat's paw' doctrine refers to the situation in which a biased subordinate, who does not have the ability to make a hiring or firing decision, is able to use the actual decision maker to trigger a negative employment action. Under the 'rubber stamp' theory, the formal decision maker mechanically consents to the discriminatory employment action at the behest of the biased subordinate. See Llampallas v. Mini-Circuits, Lab, Inc., 163 F.3d 1236, 1249 (11th Cir. 1998); Hill v. Lockheed Martin Logistics Mgmt., Inc., 354 F.3d 277, 288 (4th Cir. 2004) (en banc).

Most circuits have accepted some version of this doctrine and the Supreme Court has cited with approval the case that coined the 'cat's paw' term for this type of employment decision. Burlington Indus., Inc. v. Ellerth , 524 U.S. 742, 758 (1998) (citing Shager v. Upjohn Co. , 913 F.2d 398, 405 (7th Cir. 1990) (Posner, J.)). In Shager, the company had an unbiased committee in place to make employment decisions, but that level of formality was not enough to shield the company from liability as the Seventh Circuit found that the committee functioned as the biased supervisor's 'cat's-paw.'

The courts thus have endorsed the notion that an employer should not be able to avoid liability through willful blindness, or acting as the proverbial ostrich with its head in the sand. It is also clear that a biased employee with no final decision making power can affect an ultimate employment decision. Manufacturing performance issues, selective reporting, and other actions by a biased subordinate can all lead to an adverse employment action by an impartial decision maker based on tainted facts. However, though the 'cat's paw' doctrine is generally accepted, its scope and, specifically, the question of what degree of control a subordinate must have over the decision maker and ultimately over the adverse action, greatly differ from court to court.

There are three standards that have been used in answering those questions. The Fifth Circuit applies the most liberal test by considering whether the subordinate 'possessed leverage, or exerted influence, over the titular decision maker.' Russell v. McKinney Hosp. Venture , 235 F.3d 219, 227 (5th Cir. 2000). The Fifth Circuit and Seventh Circuit have held that 'summary judgment generally is improper where the plaintiff can show that an employee with discriminatory animus provided factual information or other input that may have affected the adverse employment action.' Dey v. Colt Constr. & Dev. Co ., 28 F.3d 1446, 1459 (7th Cir. 1994) (citing Shager , 913 F.2d at 405). Thus, providing biased 'factual information or other input' is sufficient to attach liability so long as it 'may have affected' the outcome.

The Fourth Circuit in Hill, 354 F.3d at 289, articulated a standard at the other extreme. That case held that though the subordinate does not need to be the 'formal decision maker,' he or she has to be, 'the one 'principally responsible' for, or the 'actual decision maker' behind, the action.' The Fourth Circuit declined to extend liability upon proof that the subordinate 'had a substantial influence on the ultimate decision or because he has played a role, even a significant one, in the adverse employment decision.' Id. at 291. This strict interpretation would allow an employer to evade liability with willful blindness because an employee could suffer an adverse action based solely on discriminatory bias or animus of a non-decision maker; no liability would attach because the biased individual did not have complete control over the formal decision maker. That interpretation is also the least popular; the dissent in Hill noted that the majority's decision 'puts us at odds with virtually every other circuit.' 354 F.3d at 299.

The Seventh and Tenth Circuits have required a causal link between the discriminatory animus of the non-decision maker and the adverse action. In their formulation, if a decision maker is dependent upon a biased employee for information, a mere 'paper review' of that information will not be enough to shield a company from liability. However, if the decision makers perform more than a perfunctory review and conduct their own independent research, there would be no finding of discrimination. Hence, the company can be excused even if a good portion of the information justifying the decision came from the biased employee.

A case involving that standard was scheduled to be argued before the Supreme Court last month, but was withdrawn at the last moment. Though the Tenth Circuit case will not be heard by the Court, an analysis of the facts of the case is useful in illustrating the 'cat's paw' issue. Furthermore, that case provides examples of how a firm can structure its human resources and management processes to prevent violations of Title VII.

In EEOC v. BCI Coca-Cola Bottling Co., of Los Angeles, 450 F.3d 476 (10th Cir. 2006), Peters, an African-American employee, was dismissed from his job as a merchandiser at BCI's facility in Albuquerque, NM. He reported to a Hispanic district sales manager, Grado. Peters' day-to-day supervisor was Katt, who also reported to Grado, the allegedly biased subordinate. The EEOC cited several incidents that purportedly demonstrate that Grado treated African-American employees more harshly than white and Hispanic employees.

The incident at issue involved a weekend that Peters was not scheduled to work, but due to an employee shortage, was told to come in. Grado found out that Peters was not planning to work and called a human resources supervisor, Edgar, based in Phoenix, more than 450 miles away. Edgar advised Grado to tell Peters that failure to comply with the order to report would be insubordination that could lead to dismissal. Only Edgar had the authority to dismiss an employee.

That Friday, following Edgar's instructions, Peters and Grado had a discussion about his working over the weekend and Peters' responded with, '[D]o what [you've] got to do, and I'll do what [I've] got to do.' Peters was not feeling well and went to a medical clinic on Saturday and called in sick to Katt. Katt approved his absence and then tried, but was unsuccessful, in advising Grado.

On Sunday, Pederson, the on-site human resources representative, told Edgar that Peters had a two-day suspension for an infraction two years earlier. Edgar was not told that Peters did not show up at work on that earlier occasion because he was a pallbearer in his fianc'e's son's funeral and Peters was not given permission to attend because the deceased was not his biological son. On Monday, Edgar told Pederson and Grado to terminate Peters' employment. Edgar asserts that the firing decision was based on the insubordination that Peters allegedly exhibited during his Friday discussion with Grado. It is undisputed that at the time of the termination decision neither Edgar nor Pederson knew Peters' race.

The district court dismissed the case on summary judgment because the decision maker conducted 'independent research,' that is, that there was an independent review of the personnel file to find out about the previous infraction. The Tenth Circuit reversed, finding this research insufficient and held that there was a factual issue as to whether there was a causal link between Grado's allegedly biased actions and the termination, thus warranting a trial.

Given the disagreement as to the applicable standard, the Supreme Court will likely consider a similar case in the future. In the meantime, as the BCI case teaches, companies should establish a mechanism to investigate dismissal recommendations and use independent research that is more than perfunctory. Relying on information from only one employee can be a bad idea. In BCI, the human resources manager, the decision maker, did not talk to the aggrieved employee to get his side of the story. Nor did the human resources decision maker talk to the employee's immediate supervisor to obtain more information. All the information she used in making the final decision came from the allegedly biased manager. The independent examination extended only to Peters' personnel file, not from an inquiry into the subsequent events. Arguably, the Tenth Circuit might have dismissed the claim if either of the human resource personnel had talked to the aggrieved employee or even to his immediate supervisor.

Employers have asserted that this investigative standard would carry a significant burden. Good practice, however, dictates that decision makers, to avoid the possibility of bias, should get the employee's side of a disputed situation and, to the extent possible, get input from other supervisors and managers. Moreover, the additional burden to investigate by speaking to several individuals may not be great, especially where the proposed action is termination, the most serious sanction an employer can take. It is unclear in BCI how burdensome a more thorough process would have been. Would it have been significantly more time consuming for either of the human resource personnel to speak to the employee or his other supervisor directly?

Though BCI was withdrawn, there is still the open question to which the Supreme Court granted certiorari: 'Under what circumstances is an employer liable under federal anti-discrimination laws based on a subordinate's discriminatory animus when the person who actually made the adverse employment decision admittedly harbored no discriminatory motive toward the impacted employee?' Given the disparate views of the circuits, this is an area that is ripe for clarification and one that the Supreme Court will likely address in the near future. In the meantime, by engaging in independent research, a company can help protect employees' rights while minimizing its exposure to liability stemming from a single supervisor's biased recommendation.

Debra L. Raskin is a partner at Vladeck, Waldman, Elias, & Englehard, P.C., and an adjunct faculty member at Columbia Law School. She has lectured and written on employment law matters for the New York State Bar Association and the American Law Institute/American Bar Association, among other groups. She can be reached at [email protected]. Jamie Ostrow worked as an associate in the London office of Freshfields, Bruckhaus Deringer and is currently a consulting attorney with Vladeck in New York.