Disparate Impact, Double Effect, and Euthanasia
On FindLaw today, I have a column that considers Ricci v. DeStefano, a case argued before the U.S. Supreme Court last week, in which white firefighters challenge actions taken by the City of New Haven Fire Department as violations of the Fourteenth Amendment Equal Protection Clause. The Fire Department had given an examination to determine whom to promote in its ranks but, after a hearing, decided not to certify the results of that examination because they produced a disproportionately white set of promotion-eligible individuals (and an all-white group of individuals to be automatically promoted). The Fire Department argues that if it had certified the test results, it would have been guilty of "disparate impact" race discrimination under Title VII (in which a facially neutral act resulting in a racially disproportionate outcome must satisfy "business necessity" to avoid violating the statute). My column examines the "damned if you do/damned if you don't" quality of trying to avoid both disparate treatment and disparate impact race discrimination.
In this post, I want to offer a provocative analogy to consider in thinking about disparate impact (and more specifically, in considering the Fire Department's decision to throw out the results of a test with a racially disparate impact): the case of giving a terminal patient a lethal dose of morphine. Under the laws of 48 states, physician-assisted suicide is a criminal act (the two outlier states are Oregon and Washington). This means that if a doctor deliberately gives a patient a lethal drug with the objective of ending his life, the doctor violates the law of almost every state. The law (and, to my knowledge, Catholic doctrine as well) allows, however, for a patient's pain to be treated with opiates sufficient in quantity to make the pain stop, even if that dosage is lethal to the patient. The difference between the two scenarios is the intent of the doctor: if the goal is to kill the patient, the act is prohibited, but if the goal is to relieve the pain (and has the incidental effect of also killing the patient), then the act -- given the importance of pain relief -- is permissible. The difference between intent and effect in this context tracks the distinction between purpose and knowledge (or what is sometimes called the "intending/knowing distinction").
How is this like "disparate impact" discrimination? In the following sense. When an employer uses a measure to determine how to allocate job benefits, her goal is ordinarily to award promotions only to the best employees. Once having used the measure, however, the employer learns about some of its unintended effects, and this can include a racially disparate impact. Knowing of this impact, the employer's decision to continue to rely on the test amounts to "knowing" discrimination -- that is, the employer is knowingly (if not intentionally) using a tool for awarding promotions that will favor white candidates over black candidates. If the employer knows of another, equally good, test of employee quality that does not have a racially disparate impact, moreover, the decision to continue relying on the "discriminatory" test represents a decision knowingly to engage in unnecessary discrimination.
Because the decision is not intentional, however, one who subscribes to a strong version of "double effect" reasoning might argue that such an employer's actions are permissible, despite the unnecessary and harmful discriminatory effect, just as a lethal dose of morphine is permissible, despite the lethal effect.
One important distinction, however, is that a life-threatening quantity of medication appears to be necessary to relieving many dying patients' suffering (which often includes oxygen starvation -- the sensation of suffocating -- and the terrible panic that this sensation causes), while the test in our example is apparently not necessary to awarding promotions to the most qualified employees (given the existence of another test with less racial impact). If there were two equally effective pain medicines, only one of which would kill the terminal patient, then I suspect that the law as well as Catholic doctrine would frown upon the selection of the lethal medicine over the nonlethal one. Indeed, such selection would, on its face, strongly suggest that the doctor's (and patient's) goal was in fact to end the patient's life and not merely to relieve his or her pain.
An employer's initial decision to use a particular written examination to determine whom to promote is, in all likelihood, innocent. She is not trying to discriminate; she is simply using a measure that, more or less, predicts performance quality. Once one has learned the effects of that examination, however, one must take responsibility for those effects and continue to employ the test with impunity only if it is truly necessary to rewarding the best employees with promotions. In the case of morphine and other pain-killers, we do not appear to have an equally effective (but non-lethal) way to end a suffering patient's agony. Until the same can be said of the test that the New Haven Fire Department gave its employees, it is difficult not to find a stubborn commitment to this particular test's results (and a refusal to change course) an instance of intentional discrimination. That such discrimination might be mandatory as a matter of Fourteenth Amendment Equal Protection would accordingly be unfortunate indeed.
Posted by Sherry Colb
In this post, I want to offer a provocative analogy to consider in thinking about disparate impact (and more specifically, in considering the Fire Department's decision to throw out the results of a test with a racially disparate impact): the case of giving a terminal patient a lethal dose of morphine. Under the laws of 48 states, physician-assisted suicide is a criminal act (the two outlier states are Oregon and Washington). This means that if a doctor deliberately gives a patient a lethal drug with the objective of ending his life, the doctor violates the law of almost every state. The law (and, to my knowledge, Catholic doctrine as well) allows, however, for a patient's pain to be treated with opiates sufficient in quantity to make the pain stop, even if that dosage is lethal to the patient. The difference between the two scenarios is the intent of the doctor: if the goal is to kill the patient, the act is prohibited, but if the goal is to relieve the pain (and has the incidental effect of also killing the patient), then the act -- given the importance of pain relief -- is permissible. The difference between intent and effect in this context tracks the distinction between purpose and knowledge (or what is sometimes called the "intending/knowing distinction").
How is this like "disparate impact" discrimination? In the following sense. When an employer uses a measure to determine how to allocate job benefits, her goal is ordinarily to award promotions only to the best employees. Once having used the measure, however, the employer learns about some of its unintended effects, and this can include a racially disparate impact. Knowing of this impact, the employer's decision to continue to rely on the test amounts to "knowing" discrimination -- that is, the employer is knowingly (if not intentionally) using a tool for awarding promotions that will favor white candidates over black candidates. If the employer knows of another, equally good, test of employee quality that does not have a racially disparate impact, moreover, the decision to continue relying on the "discriminatory" test represents a decision knowingly to engage in unnecessary discrimination.
Because the decision is not intentional, however, one who subscribes to a strong version of "double effect" reasoning might argue that such an employer's actions are permissible, despite the unnecessary and harmful discriminatory effect, just as a lethal dose of morphine is permissible, despite the lethal effect.
One important distinction, however, is that a life-threatening quantity of medication appears to be necessary to relieving many dying patients' suffering (which often includes oxygen starvation -- the sensation of suffocating -- and the terrible panic that this sensation causes), while the test in our example is apparently not necessary to awarding promotions to the most qualified employees (given the existence of another test with less racial impact). If there were two equally effective pain medicines, only one of which would kill the terminal patient, then I suspect that the law as well as Catholic doctrine would frown upon the selection of the lethal medicine over the nonlethal one. Indeed, such selection would, on its face, strongly suggest that the doctor's (and patient's) goal was in fact to end the patient's life and not merely to relieve his or her pain.
An employer's initial decision to use a particular written examination to determine whom to promote is, in all likelihood, innocent. She is not trying to discriminate; she is simply using a measure that, more or less, predicts performance quality. Once one has learned the effects of that examination, however, one must take responsibility for those effects and continue to employ the test with impunity only if it is truly necessary to rewarding the best employees with promotions. In the case of morphine and other pain-killers, we do not appear to have an equally effective (but non-lethal) way to end a suffering patient's agony. Until the same can be said of the test that the New Haven Fire Department gave its employees, it is difficult not to find a stubborn commitment to this particular test's results (and a refusal to change course) an instance of intentional discrimination. That such discrimination might be mandatory as a matter of Fourteenth Amendment Equal Protection would accordingly be unfortunate indeed.
Posted by Sherry Colb