Last year, in a speech that defied political gravity, President Bill Clinton threw his support behind affirmative action while coining the phrase "mend it, don't end it." Clinton's mini-couplet quickly became the mantra of those seeking to maintain the affirmative action status quo. In California, where voters will decide the future of the state's racial and gender preferences in November, opponents of the California Civil Rights Initiative have picked up on Clinton's rhetorical waffle, charging that the initiative is too blunt an instrument with which to fix the delicate apparatus of affirmative action. These programs, we are told, must be left to nimble-handed professionals, whether they be bureaucrats or lawyers.
In fact, the only hope of mending affirmative action lies in just such broad instruments as the CCRI, which tie the hands of those who would use the government to grant privileges based on race and gender. Recent reforms in the affirmative action programs in the California State Civil Service illustrate why. They also provide a template of how city and county governments in a post-CCRI California can craft programs that, while expanding opportunity, neither grant preferential treatment nor discriminate based on race, ethnicity, or sex.
Taking Executive Action
On June 1, 1996, California Gov. Pete Wilson made the first effort by any elected politician to fix affirmative action, by issuing Executive Order W-124-95. Wilson's executive order mandated that agencies and departments "eliminate all preferential treatment requirements that exceed federal or statutory law" and set goals and timetables, which are required by statutory law, based on "the employment pool possessing the necessary qualifications for the particular job classification at issue, rather than on general work force parity."
Wilson's attempt to modify the program came a quarter century into the state's affirmative action effort. In 1971, Gov. Ronald Reagan issued an executive order stating that "justice demands that every citizen consciously adopt and accentuate a personal commitment to affirmative action which will make equal opportunity a reality." In 1977, the legislature mandated that each agency and department set up an affirmative action program and "establish goals and timetables designed to overcome any identified underutilization of minorities and women in their respective workforces." The State Personnel Board was charged with overseeing the affirmative action programs.
Wilson, with the stroke of a pen, achieved what the personnel board professionals had long been unable to accomplish in the face of interest-group politics: His executive order forced the state to create an affirmative action program that, at least on paper, is based on expanding opportunity, not granting preferences.
Wilson's order had two major effects. First, although Wilson lacked the power to abolish affirmative action goals and timetables by executive action, the order mandated that state agencies conform to federal law by using "relevant labor force" data to set such goals. This meant that if the state was hiring lawyers, the benchmark labor pool would be only the people who had actually passed the state bar. Likewise, only the composition of those in the private-sector work force with a college degree could be used to determine whether minorities and women were underrepresented in state jobs that required similar credentials. Second, the order prohibited agencies and departments from using the goals and timetables to grant preferential treatment to individuals in the hiring process.
The significance of Wilson's reforms shouldn't be obscured by the welter of bureaucratic jargon: It is perhaps best demonstrated by the ferocity with which the changes were resisted.
In the late 1980s, a series of U.S. Supreme Court decisions restricted the ability of state and local governments to justify their affirmative action programs based on statistical disparities. In 1989, the Court ruled that governments must use relevant labor force data–statistics on the individuals who actually meet the qualifications for a job in the area in which the government would be reasonably expected to recruit. At the time, the state based its goals and timetables on "general labor force" data, which was simply the ethnic and gender composition of California's entire private sector work force.
The switch to using the relevant labor force is a disaster for the state's ethnic interest groups, such as Black Advocates in State Service, CAFE de California (which represents Hispanics), and the Mexican American Legal Defense Fund, who over time had secured a hiring structure that greatly benefited their constituency groups. Managers, for example, were often required to justify in writing the hiring of a white man over available minorities; managers' performance reviews were even tied to their success in meeting hiring goals. As a result of these incentives, every minority group except Hispanics was either hovering around parity or was overrepresented among state employees–even by the general labor force standard. African Americans, for example, were overrepresented on the government's payroll by nearly 90 percent. If the state were to use relevant data, the interest groups knew that their already tenuous claims to underrepresentation would become laughable, and that their power to deliver government spoils to their members would be drastically eroded.
In an effort to conform to the law, Ted Edwards, manager of the State Personnel Board's affirmative action division, called an informational hearing in 1989 to try to design a legal program. An impressive array of advocacy groups testified. Edwards says the meeting produced a "mixed reaction" and that "nothing immediately happened as a result."
Edwards called another hearing in March 1993 to implement a goal-setting process using the relevant labor force data, to bring the state into compliance with the law. "By that time, I had finished my analysis of various Supreme Court decisions, and I was convinced that some of our current affirmative action practices were not consistent with Supreme Court case law," Edwards recalls. The reaction was heated. "Advocacy organizations came in and strongly objected to our staff recommendation to change the goal-setting process to use relevant labor force comparisons," Edwards says. "It was rather blistering testimony."
As a result, the status quo held firm. Jose Perez, then assistant director for affirmative action for Caltrans, the state's transportation agency, summed up the meeting in Caltrans's 1994 affirmative action plan: "The SPB considered using relevant labor market rather than general labor force data to set affirmative action goals for employment. After considerable debate and public reaction, the five-member Board tabled the issue indefinitely."
In other words, the state chose to willfully violate the law. It wasn't until a Caltrans employee threatened to sue the state that Edwards was allowed in 1994 to create a pilot program that used relevant labor force data at Caltrans, the Department of Corrections, the California Highway Patrol, and the Department of Education.
The refusal to use relevant labor force standards, even though required by U.S. Supreme Court case law, is an object lesson in why blanket prohibitions against choosing by race, such as the CCRI, are needed if affirmative action is ever to be mended. President Clinton talks about "mending" affirmative action just as his Department of Justice works to expand such programs as 10 percent bid preferences for minorities, and there is just as little enthusiasm for "mending" affirmative action among most of its practitioners in state government.
In 1995-96, California's agencies and departments set their hiring goals using relevant labor force standards for the first time, with predictable results. Findings of disparity in the work force dropped from 26 percent to less than 5 percent. While 1,108 hiring goals were set under general labor force data in 1991-92, the last year the state compiled a report, only 235 hiring goals were set in 1995-96 after switching to relevant labor force numbers, a reduction of nearly 80 percent.
Hiring goals are set for specific classes of jobs, such as entry-level positions requiring a college degree, and often entail more than one hire. A timetable, usually between one and 10 years, is set along with the goal. For example, in 1995-96 the Department of Corrections set a goal in the Staff Services Occupational Group to hire five African Americans in two years. The department also set goals in correctional officers for 17 whites in one year, 11 Asians in one year, 13 American Indians in three years, and nine women in one year. Since goals can be set only after a finding of statistical disparity, which is assumed to be the result of discrimination, this large reduction in goals indicates that when accurately measured, discrimination against minorities plays a much smaller role in state hiring than once thought.
In fact, using relevant labor force data (RLF) made clear to department managers that two of the most underrepresented groups were whites and men. Some departments, including the Department of Education, the Franchise Tax Board, and the California Department of Health Services, even set goals for these two groups. Other departments, while finding underutilization of whites and men, chose not to set goals for them. California's Employment Development Department, for example, set eight goals for 181 hires for such groups as African Americans, Hispanics, American Indians, and women. But their own statistical analysis showed that the two most underrepresented groups in their work force were whites, who were underrepresented by 2,252 jobs, and men, who were underrepresented by 942 jobs.
The switch has been met with mixed reactions. Drew Mendelson, a spokesperson for the California State Employees Association, the state's largest union, opposes the use of relevant labor force data to set goals. "RLF may be more accurate in terms of who has individual qualifications, but it is inaccurate in that it ignores the fact that we have reached a certain balance based on inequality of opportunity before," states Mendelson. "You have to say why is it that if 10 percent of the population is black, then why aren't a tenth of the lawyers black?"
Sharon Felix-Rochon, director of equal employment opportunity at the California Department of Education, supports the move because she feels relevant labor force data provide a more accurate gauge of who is available to fill specialized jobs. "When we get to job classifications that require specific credentials, then the relevant labor pool is appropriate." Felix-Rochon, who reformed hiring in her department before Wilson's executive order, believes that the former system had degenerated. "What happened unfortunately is that there had been so much emphasis on numbers–it became a numbers game–and we failed," she says. "People would look at the numbers, and if the numbers didn't reflect the goals, we would get sanctioned. Advocacy groups would call the departments and call people racists just based on the numbers."
Lance Barnett, who became executive director of the State Personnel Board just in time to implement Wilson's reforms, also supports the move to RLF data. "It got the departments down to a workable set of goals," states Barnett. "Shifting from 'general' to 'relevant' allowed departments to focus their efforts where they could really make a difference."
Edwards, the affirmative action manager, feels that most agencies are satisfied with the new standards. "Relevant labor force is really the only thing that makes sense," states Edwards. "You can't hire in greater numbers than actually exist in the labor force."
Making California's Goals Run On Time
The second facet of Gov. Wilson's reforms addresses how departments are allowed to use their goals and timetables. Preferences, which are what the CCRI would outlaw, have to manifest themselves in a process. It is theoretically possible, as those at the personnel board uniformly claim, that goals and timetables are used merely as a census-like device to help managers examine their hiring processes and strengthen their outreach efforts.
Edwards points out that there are two components of affirmative action programs, as they have traditionally been implemented: preferential and non-preferential. Non-preferential affirmative action–or, in popular terms, "equality of opportunity"–focuses on ensuring that recruitment efforts are broad and that selection processes and criteria are free of bias. Preferential affirmative action includes programs and practices that treat people unequally in order to ensure an equality of result.
Wilson's order eliminated preferential affirmative action. Specific practices that were discontinued or prohibited include:
- Using separate hiring lists for hiring individuals from protected groups;
- Requiring managers to write letters justifying the hiring of a white person when there were individuals from protected groups on the short list;
- Allowing a department's affirmative action officer to unilaterally veto a hire;
- Evaluating managers, in part, on how well they meet their affirmative action goals; and
- Requiring that either a woman or a minority sit on every hiring panel.
Barnett, the personnel board chief, maintains that the focus under the new system is on ensuring that the process is fair. "The key is to make sure that your qualified candidate pool resembles the relevant labor force and then rely on your selection criteria to be unbiased to get proportional representation," he says.
At least on paper, it appears that the board has succeeded in crafting an affirmative action program that doesn't rely on preferences. As such, it provides an example of how affirmative action will look in post-CCRI California. "The E.O. has accomplished in the state service what the CCRI will accomplish more broadly, which is taking preferences out of our employment processes," explains Ted Edwards, who nevertheless opposes the CCRI for not giving government officials enough flexibility. "The CCRI does not in and of itself eliminate affirmative action. Affirmative action is broader than just preferential actions."
Twofer White Males?
Even Ward Connerly, the outspoken chairman of the CCRI campaign, voices support for the parameters of the state personnel board's new program. "The California Civil Rights Initiative is designed to end affirmative action programs that grant preferences based on race, ethnicity, and sex, but we really want government agencies to examine their hiring practices to make sure they do not discriminate," states Connerly. "What the state has done under Governor Wilson's leadership is a clear example of what can be done in a post-CCRI era."
Some experts, however, feel that quantifying goals by race, ethnicity, and gender is preferential and unconstitutional. Indeed, Wilson, having accomplished all he could by executive action, is suing the state, arguing that the statutes that require goals and timetables violate the U.S. Constitution. "The goal is by definition discrimination based on race and gender," states Robert Corey, an attorney at the Pacific Legal Foundation who is litigating the governor's case. "Its very existence gives managers license to discriminate against innocent people based on the immutable characteristics of race and gender."
There is often a disconnect between official policies, especially those set by overseeing agencies, and how things actually work in practice. Indeed, a memo titled "Revision to Fiscal Year 1995-96 Affirmative Action Goals and Timetables," from the Department of Alcohol and Drug Programs, supports Corey's contention and indicates that civil service hiring may not be as race- and gender-blind as those in power would have us believe.
In this memo dated August 15, 1995, Lisa Fien, the department's equal employment opportunity officer, wrote: "Eight vacancies are anticipated to be filled in the Staff Services Occupational Grouping [the entry level civil service position for college graduates]. Four (4) vacancies will be filled by whites; four (4) filled by men." Note that these are not only quotas, they are quotas set for whites and men, in which a white man can be a "twofer" hire, a bonus formerly reserved for minority women. Nevertheless, such quotas are supposed to be a thing of the distant past.
While the personnel board is not taking a position on Wilson's case, Edwards defends the use of goals and timetables as a way to let department heads know what is going on in their work force. "Our responsibility is to ensure that employment practices are fair to all groups," says Edwards. "Unless you monitor your employment practices with statistical tools, to see that equal employment opportunity is a reality, it may not be." Edwards also points out that at least under the new process, it is the affirmative action office, not the hiring managers, that tracks the goals and timetables.
This final debate will be resolved only in court. But the state's affirmative action program has already been retooled from what had degenerated into an often irrational ethnic spoils system. Insofar as President Clinton and other affirmative action advocates claim to support only those programs that focus on equality of opportunity, as opposed to a quota-based equality of result, then they should lend their support to the CCRI. After all, it is only through such efforts that affirmative action has ever been mended.
Michael Lynch (email@example.com) is a public policy fellow at the Pacific Research Institute in San Francisco.
Start your day with Reason. Get a daily brief of the most important stories and trends every weekday morning when you subscribe to Reason Roundup.