Implicit Biases: A Hypothetical
Building on the last post that decried ineffectiveness of the perpetrator model of racial discrimination, this post will buttress the claim that well-meaning individuals can act in ways that hinder the goals of equal opportunity and that they often do so without the knowledge that their actions are caused by racial stereotypes. Linda Krieger asserts that “the entire normative structure of anti-discrimination law–is based on an assumption that decision-makers possess ‘transparency of mind,’ or that they can accurately identify why they are about to make, or have already made, a particular decision.” Using a hypothetical, I will show that this structure is a clumsy tool for assessing the discrimination liability of companies.
To explain, I will refer to a hypothetical decision-maker, Jeff. Jeff is a white male raised in the United States. He does not hold any racist or sexist beliefs on which he would consciously act. Let us further assume that Jeff is the manager of assembly line operations at an automobile assembly plant that employs at least one practice that has an adverse impact on women or minorities. Part of Jeff’s job is to promote line workers to the position of line supervisor in the different divisions of the plant. The line supervisor position requires knowledge of the responsibilities of the various assembly line workers in the division, attention to detail, the ability to encourage workers to produce a high-quality product, and the willingness to hold underproducing workers to account for failure to perform. Jeff evaluates the employees who are up for the promotion with an exam that tests their knowledge of the plant and a personality exam that seeks to quantify the test-taker’s leadership skills. Employees who achieve a certain benchmark score on both tests have interviews with Jeff who then makes the final decision. Although women and men both tend to score very well on both tests, men receive the promotion twice as often as women with similar scores. Additionally, although seventy percent of the women who have been interviewed are Asian-American, there has never been an Asian-American woman in the position of line supervisor. The women complain that Jeff thinks they are timid and not assertive enough to apply the appropriate pressure to underperforming workers. Jeff does not believe that he is consciously overlooking female or Asian-American candidates. Instead, he explains that the specific Asian-American females who have come up for the promotions simply haven’t got “the right stuff”.
The Asian-American women don’t have a discrimination claim under the perpetrator model. There isn’t any evidence that Jeff is intentionally dismissing the Asian-American candidates. The plant that employs them will counter with the fact that Jeff is a fair-minded decision-maker with years of experience. The company does not have another reliable method of choosing the most qualified candidates from among the pool of high-scoring applicants. Jeff’s interviews are able to draw out information that can’t be elicited in a battery of tests and his decisions have stood the company in good stead for years. As such, the practice of allowing Jeff to conduct interviews as the final phase in the promotional process is related to legitimate employment goals and legal.
The preceding hypothetical presents two important points which are often outcomes of these situations: one, the auto assembly plant has acted to protect its interests, its reputation and its supervisor, Jeff, who is acting in good-faith. Two, the burden of a practice with discriminatory results lies on the innocent plaintiffs, the Asian-American women.
Jeff may be implicitly associating the Asian-American women he interviews with weakness or submissive personality traits without being consciously aware of the association. Although he doesn’t know it is there, it influences his decision-making and acts as a barrier to the career advancement of Asian-American women in the plant. Researchers at Yale University and the University of Washington have developed a series of implicit association tests. The tests originated as tools to assist the researchers in delving into “the unconscious roots of thinking and feeling”. On their website, individuals may take an implicit association test to learn more about their own unconscious preferences and beliefs. One of the tests gauges the implicit associations and preferences the test-taker has for White Americans and African-Americans.
The entire test is a grouping exercise and must be done quickly so that the subject acts on her first impulse. Test takers are presented with computer generated faces that are stereotypically White- or Black-American which the test-taker must group under the White-American or Black-American heading. The faces are not ambiguous and clearly belong in one category or the other. Next, the subject must catalog words under “good” or “bad”. The words are also unambiguous, such as “wonderful” or “horrible.” The next phase requires test-takers to sort “White-American” and “good” on one side of the screen and “Black-American” and “bad” on the other side of the screen. The same faces and words will pop up on the screen and the subject must group them as quickly as possible. The final phase switches the categories with “Black-American” and “good” on one side and “White-American” and “bad” on the other. The final two phases occur in random order and do not materially affect the final score. The test interprets the subject’s response speed in the final phases. A subject has “automatic preference for Black” if she responds faster in the phase of linking African-American faces with pleasant words and “automatic preference for White” if she responds faster to linking white faces and pleasant words. The preferences vary from strong or moderate, to little or no preference.
The implicit association tests do not seek to smoke out prejudiced people who approve of discrimination. Instead, they attempt to dig into the private attitudes that individuals are unwilling to express and the private attitudes that may be hidden from conscious thought and feeling. The researchers have found that people with an automatic preference for White are able to avoid discriminatory behavior by making conscious efforts to check the preference. By revealing the preferences and racial associations that may lie beneath the surface, the implicit association race test has interesting implications for antidiscrimination law. Jeff could be just the sort of person who, without harboring any conscious negative feelings toward African-Americans, would score a strong automatic preference for Whites on this test. Though he may not be aware of it, his difficulty in being able to associate good things with African-Americans and the ease with which he associates positive attributes to White-Americans may be having deleterious effects on the diverse workforce he oversees.
Combating unconscious discrimination requires fewer recriminations and more tough conversations, less self-defense and more self-reflection. We can make progress in race relations, but progress will require a radical re-thinking of our basic assumptions about racism. Just as the Civil Rights Movement swept away entrenched ideas about racial superiority and inferiority, completion of the Movement demands that we remain open to new ways of confronting racism. This time, the battle lines may not be drawn in black and white. In the second wave of the Movement, we all live in shades of gray.
To explain, I will refer to a hypothetical decision-maker, Jeff. Jeff is a white male raised in the United States. He does not hold any racist or sexist beliefs on which he would consciously act. Let us further assume that Jeff is the manager of assembly line operations at an automobile assembly plant that employs at least one practice that has an adverse impact on women or minorities. Part of Jeff’s job is to promote line workers to the position of line supervisor in the different divisions of the plant. The line supervisor position requires knowledge of the responsibilities of the various assembly line workers in the division, attention to detail, the ability to encourage workers to produce a high-quality product, and the willingness to hold underproducing workers to account for failure to perform. Jeff evaluates the employees who are up for the promotion with an exam that tests their knowledge of the plant and a personality exam that seeks to quantify the test-taker’s leadership skills. Employees who achieve a certain benchmark score on both tests have interviews with Jeff who then makes the final decision. Although women and men both tend to score very well on both tests, men receive the promotion twice as often as women with similar scores. Additionally, although seventy percent of the women who have been interviewed are Asian-American, there has never been an Asian-American woman in the position of line supervisor. The women complain that Jeff thinks they are timid and not assertive enough to apply the appropriate pressure to underperforming workers. Jeff does not believe that he is consciously overlooking female or Asian-American candidates. Instead, he explains that the specific Asian-American females who have come up for the promotions simply haven’t got “the right stuff”.
The Asian-American women don’t have a discrimination claim under the perpetrator model. There isn’t any evidence that Jeff is intentionally dismissing the Asian-American candidates. The plant that employs them will counter with the fact that Jeff is a fair-minded decision-maker with years of experience. The company does not have another reliable method of choosing the most qualified candidates from among the pool of high-scoring applicants. Jeff’s interviews are able to draw out information that can’t be elicited in a battery of tests and his decisions have stood the company in good stead for years. As such, the practice of allowing Jeff to conduct interviews as the final phase in the promotional process is related to legitimate employment goals and legal.
The preceding hypothetical presents two important points which are often outcomes of these situations: one, the auto assembly plant has acted to protect its interests, its reputation and its supervisor, Jeff, who is acting in good-faith. Two, the burden of a practice with discriminatory results lies on the innocent plaintiffs, the Asian-American women.
Jeff may be implicitly associating the Asian-American women he interviews with weakness or submissive personality traits without being consciously aware of the association. Although he doesn’t know it is there, it influences his decision-making and acts as a barrier to the career advancement of Asian-American women in the plant. Researchers at Yale University and the University of Washington have developed a series of implicit association tests. The tests originated as tools to assist the researchers in delving into “the unconscious roots of thinking and feeling”. On their website, individuals may take an implicit association test to learn more about their own unconscious preferences and beliefs. One of the tests gauges the implicit associations and preferences the test-taker has for White Americans and African-Americans.
The entire test is a grouping exercise and must be done quickly so that the subject acts on her first impulse. Test takers are presented with computer generated faces that are stereotypically White- or Black-American which the test-taker must group under the White-American or Black-American heading. The faces are not ambiguous and clearly belong in one category or the other. Next, the subject must catalog words under “good” or “bad”. The words are also unambiguous, such as “wonderful” or “horrible.” The next phase requires test-takers to sort “White-American” and “good” on one side of the screen and “Black-American” and “bad” on the other side of the screen. The same faces and words will pop up on the screen and the subject must group them as quickly as possible. The final phase switches the categories with “Black-American” and “good” on one side and “White-American” and “bad” on the other. The final two phases occur in random order and do not materially affect the final score. The test interprets the subject’s response speed in the final phases. A subject has “automatic preference for Black” if she responds faster in the phase of linking African-American faces with pleasant words and “automatic preference for White” if she responds faster to linking white faces and pleasant words. The preferences vary from strong or moderate, to little or no preference.
The implicit association tests do not seek to smoke out prejudiced people who approve of discrimination. Instead, they attempt to dig into the private attitudes that individuals are unwilling to express and the private attitudes that may be hidden from conscious thought and feeling. The researchers have found that people with an automatic preference for White are able to avoid discriminatory behavior by making conscious efforts to check the preference. By revealing the preferences and racial associations that may lie beneath the surface, the implicit association race test has interesting implications for antidiscrimination law. Jeff could be just the sort of person who, without harboring any conscious negative feelings toward African-Americans, would score a strong automatic preference for Whites on this test. Though he may not be aware of it, his difficulty in being able to associate good things with African-Americans and the ease with which he associates positive attributes to White-Americans may be having deleterious effects on the diverse workforce he oversees.
Combating unconscious discrimination requires fewer recriminations and more tough conversations, less self-defense and more self-reflection. We can make progress in race relations, but progress will require a radical re-thinking of our basic assumptions about racism. Just as the Civil Rights Movement swept away entrenched ideas about racial superiority and inferiority, completion of the Movement demands that we remain open to new ways of confronting racism. This time, the battle lines may not be drawn in black and white. In the second wave of the Movement, we all live in shades of gray.