Why Race-Based Data Matters
Stay up to date with the latest headlines via email.
EDITOR'S NOTE: This is the third in a series of three articles by Sally Lehrman, a freelance medical and science writer and Expert Fellow of the Institute for Justice and Journalism at the University of Southern California's Annenberg School for Communication.
Parents in Missoula, Montana, demanded that teachers learn more about Native American culture. Middle school instructors in San Francisco began to revamp their discipline policies. Students in Los Angeles organized to delay the state high school exit exam.
Using data on teacher demographics, expulsion records, test scores, and other measures, these groups identified what they saw as patterns of structural racism that led to disengagement by students of color and ultimately, fewer opportunities for higher education. "We believe institutional racism is the single largest barrier to achievement by students of color," says Alexa Hauser, who coordinates the Principals' Anti-Racist Leadership Institute in the San Francisco Bay Area.
Education is just the latest arena in which community leaders, attorneys, and social scientists have increased their reliance on race-based data to identify hidden inequalities and search for ways to address them. Civil rights specialists have demanded traffic stop information to check for racial profiling. Academic researchers have pored over job statistics to hunt for disparate treatment. And slowly, courts are beginning to respond to the potential of subtle bias in the workings of organizations that seemed open to all. Even as overt discrimination wanes in American society, the new research suggests, racism has persisted in more subtle ways, often because of policies and practices built long ago into our institutions. By focusing on embedded patterns of disparity instead of hunting for bigotry, these activists hope Americans can be more effective in rooting out racism.
"My hope is that we will stop looking for the bad actor - the "evil racist" - and look at the way all our institutions work on a foundation of stereotypes and biases," says civil rights attorney and Stanford University professor Michelle Alexander. "People have become increasingly aware of the need to collect this data to identify racial disparities and use it to solve them."
School administrators and parent activists say that race-based analysis has helped them uncover stereotypes and institutional bias in everything from hiring practices and teacher placement to curriculum materials and student discipline. In effect, they say, many schools operate under a multi-faceted "tracking" process that directs some racial minorities right out of the educational system.
In the San Francisco Unified School District, for example, around 5 percent of Latinos and 4 percent of black students were enrolled in gifted or advanced placement courses in 1999, compared to 23 percent of whites. In contrast, 9 percent of African American children were suspended or expelled that year, while only 2.4 percent of whites were unable to stay in school. Yet teachers have little idea of their own differential behavior, according to Hauser of the principals' initiative. "It's pretty subconscious for most people," she says. "When it's pointed out to them, they're shocked."
While people generally have become more comfortable cross-culturally, assumptions and stereotypes still influence our perceptions of one another and limit opportunities for ethnic minorities, some scholars report. In one test of this type of unconscious bias, Stanford University linguist John Baugh measured how San Francisco Bay Area apartment renters responded to inquiries made using African American, Mexican American and Standard American English dialects. He found that African American and Mexican American voices had a much harder time making appointments in predominantly white neighborhoods.
The same subconscious bias that tracks students through school and renters into certain neighborhoods can also track people through the rest of life, according to civil rights activists. In a project on law enforcement led by Alexander, for instance, the Northern California American Civil Liberties Union analyzed highway stops in 2000. They found that Latinos were three times as likely as whites to be pulled over and searched by officers - even though they also were the group most often let go without a citation. Confronted with the data, the California Highway Patrol agreed to stop searching cars for drugs if there was no evidence of criminal activity.
Alexander says the assumptions that contribute to disproportionate highway stops lead to differential treatment at every step of the criminal justice system. The impact grows like a pyramid scheme, helping to create a cumulative overrepresentation of African American and Latino men behind bars. From juvenile arrests on up, stereotypes can influence who is stopped, what the charge is, who is offered a plea bargain, whom the jury believes and who is slapped with a harsh sentence or given a chance to reform, she suggests.
A number of studies have pinpointed disparities in sentencing. In Georgia between 1990 and 1995, for instance, African Americans eligible for life in prison for drug offenses were five times as likely to get that punishment than whites. In Philadelphia, African American men were 38 percent more likely than others to receive the death penalty between 1983 and 1993, according to an analysis that controlled for crime severity and criminal history.
Racism or rationality?
When 60 law enforcement agencies across the state began voluntarily collecting similar information, some were surprised by the results, says Alexander. "Many said there were disparities they did not know how to explain," she says. "That has forced some introspection among law enforcement agencies that has been incredibly productive."
Some insist that the disparities make sense, however. In Sacramento, the police department learned that African American and Hispanic drivers were nearly twice as likely as whites to be detained at least 30 minutes and searched after routine traffic stops. Yet contraband was found in the cars of all three racial groups at the same rate. Howard Greenwald, a University of Southern California policy professor who analyzed the data, said the differences might be justified because blacks represented 40 percent of reported suspects, 40 percent of people on probation, and almost 45 percent of parolees in the area.
The San Francisco Police Department has been compiling traffic stop data by race, gender and age for six months. "I don't believe personally this department has racially profiled," says Alex Fagan, chief of police. "The ACLU has looked at departments and found many of them do stop disparate to the population - and that may be true here. But what does that mean? I don't know. It may not be racism."
Discrimination on the job
Fagan and Greenwald do agree, however, that collecting the data is a useful way for law enforcement to monitor potential inequities. Such efforts follow in the footsteps of housing and employment agencies, where such information has long been helpful in identifying discrimination. Even in these arenas, however, researchers are now finding that a more comprehensive analysis sheds new light on old problems. Rutgers law professors Alfred and Ruth Blumrosen, who both helped establish the Equal Employment Opportunity Commission in 1965, spent four years studying data filed by employers to the agency in 1999. Using a very conservative statistical method, they looked for unfair treatment that had only a 5 percent probability of happening by chance. Still they estimated that 2 million minorities and women across the country had encountered job discrimination that year.
In California, African Americans, Asian Americans and Native Americans faced discrimination once every four times they sought an employment opportunity, the Blumrosens found. Women encountered it more than a fifth of the time. The couple dug deeper and identified 1,400 companies they described as "hard-core" discriminators, where there was only one chance in 100 that their practices occurred by accident and which had maintained this pattern for more than a decade. They also discovered that some industries could be counted on for unfair treatment: the top three with apparent biases against minorities were hospitals, department stores, and eating and drinking places. "When you look at it through the lens of our statistics, it becomes almost a manageable problem," says Alfred Blumrosen. "You can see where you could proceed to reduce the number of affected workers over time."
The Blumrosens' statistical method focuses in on discrimination above and beyond what might be occurring through unconscious prejudices or societal forces, they say. "They're so far below the others, they stick out like a sore thumb," Ruth Blumrosen points out. The couple chafes at the Equal Employment Opportunity Commission's tepid response to their findings, however: a paltry half dozen investigations. Even though Title VII of the Civil Rights Act of 1964 covers unintentional discrimination if plaintiffs can show a pattern of disparate treatment, the agency has a policy against pursuing cases without specific complaints.
Changes in the courts
Likewise, the vast majority of courts have required proof of direct discrimination, such as bigoted remarks or racial slurs, legal experts say. Only recently have theories about inter-group bias and institutional racism begun to seep into court arguments, according to Linda Hamilton Krieger, a law professor at the University of California School of Law in Berkeley. In a recent suit by six California women against giant Wal-Mart Stores Inc., plaintiffs complained that the company culture leads to sex discrimination. "Where decision-making is subjective and unstructured, combined with a tendency toward stereotyping, you end up with systematic discrimination," Krieger explains.
A case in 1989 against accounting firm Price Waterhouse helped make judges more open to such a perspective, she says. A district court decided that the global company had discriminated against Ann Hopkins when it refused to promote the senior manager because, partners had worried, she was abrasive and too masculine. The U.S. Supreme Court agreed that gender seemed to have entered unfairly into the decision, but said Price Waterhouse should have the chance to show that it would have made the same decision anyway.
"The law is not developed to reckon with this type of discrimination," Krieger says. As overt discrimination is fading, however, courts are gradually changing in response, she and other legal experts report. Assumptions about capabilities, passivity or assertiveness, and the reasons for success or failure just of few of the many ways that bias can subtly influence worker opportunities, according to Krieger. The new understanding of unconscious stereotyping has combined with increasing attention to data showing patterns of disparate outcomes.
"People are starting to see all this evidence of enduring disparity in treatment even when you control for all of the variables that might affect it," Krieger says. "I call it 'un-attentional' bias. It's very easy to think you're making a decision on legitimate, nondiscriminatory thinking, but you're operating on stereotype."
Bias in Employment
Bias in the Courts/Criminal Justice System
Bias in the Educational System
For the Scientifically-Minded
Education & Criminal Justice Organizations
The Institute for Justice and Journalism was created at the University of Southern California's Annenberg School for Communication with Ford Foundation funding to strengthen news coverage and public understanding of justice and civil rights issues.
Resources compiled by Shannon Seibert.