WASHINGTON -- Black federal employees were more than twice as likely to be fired as their white, Hispanic or Asian counterparts, according to a report made public yesterday by the Office of Personnel Management.
The disparity was seen regardless of occupational category, pay level, education, agency, geographic location, age, performance rating, seniority or attendance record.
"The racial gap persists within every category of risk we assessed," Dr. Hilary Silver, a sociologist from Brown University who conducted the study for the agency, said in a telephone interview yesterday.
The study was begun in 1994 after the Office of Personnel Management discovered what it called "a disturbing pattern" of dismissals showing that minority employees were discharged at a higher rate than white employees. The Clinton administration and members of Congress were concerned about the disparity in dismissals and urged a government-wide investigation.
"These numbers were like a smoke detector," said James B. King, the office's director, at the time. "They signaled the need for immediate action."
At the request of the office, Dr. Silver looked at 10,544 federal workers who were involuntarily dismissed during the 1992 fiscal year. Nearly 12,000 federal workers were discharged that fiscal year, 52 percent of whom were black.
Blacks make up 28 percent of the federal civilian work force of 2.2 million, excluding postal workers. Dismissals of blacks occurred at a rate of 10 per 1,000 employees, compared with three per 1,000 among other workers.
But race, Dr. Silver said, was only one of the top five predictors of the the probability of a person's being fired. She stopped short of concluding that higher dismissal rates of minority members could be traced to discrimination, saying only: "We should proceed as if there might be racial bias involved."
Focus groups and other agency reviews included in the study, she suggested, offered possible explanations: bias or lack of cultural awareness, poorly trained supervisors and managers; or a general inability on the part of minority employees to work "the old boy" network.
"There was the general belief that minorities didn't have the connections or didn't know how to work 'the system' to keep from getting fired," said Dr. Silver.
Besides race, other important predictors of firing were holding low-level jobs with low pay, poor job performance ratings, a lack of recognition or promotion, suspension and being male.
Women, college graduates, white-collar workers and those with higher pay or working outside a metropolitan area were least likely to be fired, she said.
The firings studied were primarily for "misconduct," a category that besides poor performance includes theft, embezzlement and insubordinate acts like striking or throwing something at a supervisor. When dismissals were based solely on poor job performance, about one-third of the 1992 firings, there was no racial differential, Dr. Silver said.
The disparity in dismissal rates did not affect all minority groups equally. The rates for African-Americans and Native Americans were significantly higher than others. Hispanics were dismissed at a rate slightly higher than non-Hispanic whites. Asian and Pacific Islanders are discharged at the same rate as nonminorities.
African-Americans who were entitled to appeal disciplinary actions were 2.4 times more likely to be dismissed, while full-time employees who lack a right to appeal were 3 times at increased risk.
The study was begun before the Clinton administration undertook its review of affirmative action programs that give preferential hiring and promotions to women and minorities.