Affirmative action works.

A conservative speaks out

You will, of course, wish to know my credentials for presenting a conservative case for affirmative action in employment. Well, first, as a social scientist with the requisite degrees and academic publications, I am licensed to diagnose the ills of the American workplace from my seat in the ivory tower. But I am also a management consultant with clients who often call on me to prepare their cases in fair-employment disputes, testify as an expert witness, and provide other services adversarial to plaintiffs claiming they have suffered discrimination. My politics are conservative and I have a strong Republican voting record. I live in Alabama, a state that takes a dim view of social engineering. I hold a professorship in the College of Business at Auburn University, not exactly a hotbed of liberal thinking. All this may give me an insider advantage in disputing the conventional wisdom of the Right, but my actual views on the issue at hand probably won't bring me an invitation to work on the civil rights platform plank at the next Republican convention.

Apart from abortion, affirmative action is arguably the most loaded political issue of the day, and the least rationally argued. What passes for debate is mainly the clash of opposing evangelists with messages full of sound bites, catch phrases, and code words preaching disinformation to choirs of true believers. One choir gets the message: "Because you are black or female you are a victim and we are going to make them give you a job." The other choir hears: "They want to take away your job and give it to a black, or a female, or a black female." My effort will be to examine the facts and the arguments on the way to my conclusion, which is that affirmative action works rather well, at least in the context of employment. The points I will make apply less, if at all, to other race-conscious initiatives that carry this label.

Let's begin with the metaphor of the playing field. One argument is that the field was leveled in 1964 with the passage of Title VII of the Civil Rights Act. Or that, with more than thirty years of enforcement, it must be level by now. Well, it really is more level than it was. The mainly liberal view that equal employment opportunity somehow ended with the presidency of Ronald Reagan and that the Republicans spent twelve years turning back the clock is simply wrong. Republican presidents since Richard Nixon tended to be much more progressive on this issue than they were ever able to admit to their followers on the right. Most overt discrimination against blacks has ended. Systemic or unintentional discrimination resulting from seemingly neutral hiring practices, such as employment tests that commonly had an adverse effect on black applicants, were addressed by the 1971 Supreme Court decision in Griggs v. Duke Power Company.

Nevertheless, minorities and females continue to be underrepresented in many types of jobs and concentrated in others. For example, most of a company's accountants may be white and most of its custodians black. That may be because the company recruits for entry-level accounting jobs at colleges and universities where enrollment is predominantly white. Perhaps the company relies on employee referrals (the old-boy network). Another explanation has to do with how decisions are made when a pool of applicants includes both blacks and whites and it is necessary to select those who are "best qualified." A hiring decision is ultimately a judgment call made after a job interview by managers and supervisors with preconceived ideas about the attributes a candidate should have attributes usually similar to their own. Although job interviews are a well-established management prerogative, research has shown that they are notoriously unreliable predictors of successful choices. Even at entry level, most individuals who make hiring decisions are white males; the higher the job level, the more this is likely to be the case. In the typical "multiple hurdle" hiring process, all the applicants who make it to the final pool are "qualified," and this group often includes minorities and women. Some do get hired, of course, and probably some are selected because they are black or female. In the aggregate, however, hiring officials tend to follow their instincts, and even in the absence of ill will or bias this works to the disadvantage of minorities and women. I have observed this pattern in my consulting practice, which, I should add, is not limited to the South. There are other factors that help to explain the underrepresentation of minorities and women in certain jobs, and sorting them out would be very difficult. But to say that discrimination is not part of the problem is about as credible as saying that it explains everything.

There are, of course, remedies for discrimination. An applicant who has been discriminated against can file a charge with the Equal Employment Opportunity Commission (EEOC). Even if the EEOC finds for the complainant, however, the agency cannot enforce its ruling but must take the case to federal court. Very few such cases are actually litigated by the EEOC; usually, the charging party is given a private right to sue. But the process is expensive and proving intentional discrimination is difficult. The employer need only state a "legitimate, nondiscriminatory reason" for the decision - for example, that a black applicant for an accountant position had less experience than a white applicant. Once the plaintiff would have been able to prevail - and to receive appropriate relief, including a job placement, back pay, and possible punitive and compensatory damages by showing that the white applicant did not have more experience. But since the Supreme Court decision in St. Mary's Honor Center v. Hicks (1993), the plaintiff has been required to prove that the reason was a "pretext for discrimination"; he or she must uncover a prejudicial statement made by a hiring official, or a similar "smoking gun." That might have happened thirty years ago but today most employers are sophisticated enough to avoid doing or saying things that would expose them to legal risk.

But does this justify laws and regulations requiring employers to hire people because of their race or gender to meet quotas, thereby bringing about reverse discrimination - as was suggested in TV spots used by the Jesse Helms campaign in the 1996 Senate campaign in North Carolina? That does seem unfair, and apparently enough voters agreed to re-elect Senator Helms. Maybe such laws and regulations should be repealed outright, or, as President Bill Clinton has suggested, at least be fixed.

So let's look, beginning with Title VII of the Civil Rights Act as amended. The part on racial quotas is in 703(j), which says: "[No employer is required] to grant preferential treatment to any individual or group because of the race, color, religion, sex, or national origin of such individual or group on account of an imbalance that may exist...." And legislation passed in 1991 discourages employers from using any such practices. No quotas here. So perhaps the problem rises not out of statutes but from regulations. Presidential Executive Order 11246, issued by Lyndon Johnson in 1965, established the Office of Federal Contract Compliance Programs, which requires employers with government contracts a category that includes most major corporations as well as many small businesses, banks and other financial institutions, and colleges and universities - to have affirmative-action programs. That the order originated with LBJ definitely raises a flag for conservatives. But it was also signed by Richard Nixon, Gerald Ford, Ronald Reagan, and George Bush, who was so adamant about not having quotas one had only to read his lips.

Did these presidents fail to read what they were signing? Or did they actually read the order and decide it was reasonable? Perhaps they conferred with employer groups - overwhelmingly conservative and Republican - and learned that the business community was generally supportive. That, in fact, is how it happened. Furthermore, the order does not require preferential treatment or quotas. Rather, it requires covered employers to compare utilization of minorities and women in various job groups in their work force with the availability of qualified minorities and women in the relevant labor market. The key words here are "availability" and "qualified." That 35 percent of a given population is black does not mean that all of the blacks are qualified to be accountants. Availability would consider the percentage of blacks in the population but focus mainly on the component of that percentage qualified for a particular job. Where underutilization is found, employers are required to set goals and "use all reasonable efforts" to hire qualified minorities and women. Employers are allowed to establish these goals based on their own determination of availability.

Assume, for example, that a federal contractor has fifty accountants, of whom only two, or 4 percent, are black. The employer analyzes the relevant labor market and determines that the availability of qualified blacks is 10 percent. If ten accountants are to be hired this year, the employer would set a goal equal to availability, and attempt to hire one black applicant. In subsequent years, three more black accountants will be hired and underutilization will no longer exist in this job category with this employer. At this juncture, devotees of the level-playing-field metaphor might again make their point: that if affirmative action was ever necessary, it isn't necessary now. They're right - and that very same executive order anticipates their argument. Once the goal has been reached, the employer is no longer required - in fact, is not allowed - to set goals for hiring black accountants.

Some employers with no government contracts nevertheless adopt voluntary affirmative action plans. Underutilization of women and minorities exposes an employer to legal liability under the disparate impact theory of discrimination established by the Griggs decision. If a group of plaintiffs can identify hiring practices which appear to screen out blacks at a disproportionate rate, the burden is on the employer to justify those practices. In 1989, the Supreme Court decision in Wards Cove Packing Co. v. Antonio modified the case law, making it somewhat more difficult for plaintiffs to sue under the disparate impact theory, but Congress then enacted Griggs into the Civil Rights Act of 1991, signed by President Bush.

Doesn't this open the door to potential abuse? What about employers fearful of getting sued and therefore rushing to get their numbers up? There are probably some private employers who are sympathetic toward minorities and who want to be socially responsible. And, of course, the public sector is full of bleeding heart liberals. Won't both the supercautious and the hyperliberal go overboard in the direction of reverse discrimination?

No doubt some do, but such informal, ad hoc affirmative action is recognized and forbidden in case law. In Daugherty v. Barry, for example, a U.S. District Court ruled that the District of Columbia violated Title VII when eight eligible white applicants were bypassed in favor of two black applicants. The court found that the hiring decision was based on the city administrator's "personal vision" rather than a properly set goal. Not only case law but also federal regulations covering voluntary affirmative-action plans establish procedural safeguards - prohibiting, for example, laying off whites to maintain a racial balance or refusing to hire qualified white males - provisions very similar to those applying to government contractors.

Finally, in particular cases federal courts may require employers to adopt affirmative-action plans until they achieve compliance. In 1971 the Alabama Department of Public Safety had not been able to find even one black applicant qualified to be a state trooper, a problem that a federal judge solved by imposing an affirmative-action plan under which the judge determined availability, established goals, and instructed the department not merely to "use all reasonable efforts" but to "find and hire" qualified black applicants. Court-ordered plans imposed when discrimination is found to be pervasive and egregious can be much more stringent than voluntary plans or those required by executive order, but they are still designed to end when goals are met. Such cases are rare today, and most of the early orders have been vacated.

With the possible exception of this last category, it will be seen that affirmative action amounts to little more than a mildly proactive approach to equal employment opportunity. It most certainly does not impose gender and racial preferences or quotas, as its opponents would have us believe, nor does it go as far as its supporters would probably like. True, it is obviously not color-blind or gender-blind. Moreover, a basic tenet of conservative orthodoxy warns against being even mildly proactive; in the neoclassical school of economics (the one we attended), market forces are considered the appropriate means of dealing with social questions.

But the doctrine of market efficacy assumes free mobility of resources: Capital (equity or debt funding) flows to the enterprise where the return is highest, the latest technology is used, and human resources are deployed where their skills match the tasks to be performed. The underrepresentation of qualified minorities and women in certain jobs reveals a barrier to the mobility of human resources. Some neoclassicists would argue that an imperfect market is better than a governmentally regulated market; firms that do not hire the best-qualified workers of any race or sex will suffer, much as they would if they acquired too much debt load or chose the wrong accounting software. In the long run it all works out. But, as John Maynard Keynes (admittedly not a neoclassicist) once pointed out, in the long run we are all dead. Moreover, discrimination in hiring imposes costs not only on individuals but on society; if human resources are not put to their highest and best use, the economy performs less well and expenditures for social services and income maintenance rise; and this violates another neoclassicist assumption, which holds that in a true market economy enterprises pay all the costs of production and do not shift them to society. To the extent that affirmative action matches up minorities and women with jobs for which they are qualified and are as likely as not to be "best qualified," the cost to the enterprise (and to the economy) becomes much smaller and in some cases disappears entirely. Perhaps conservatives should also ponder the possibility that that black accountant may find a better software package for the company's financial information system, receive a generous increase in salary, and become a Republican favoring more and better tax breaks for corporations.

More basically: As a conservative I recognize an obligation to support programs that promote personal responsibility and join opportunity with merit. That sounds a lot like what those people who conceived the "Contract with America" had in mind. If we expect people to buy into the work ethic, we might want to take steps to ensure that that attitude is rewarded. It is not good for an accountant always to be called and never to be chosen. It is not good for a welfare mother who has just completed a clerical-skills training program not to be able to find a job as a secretary. It is not good for the country when qualified women and minorities are underrepresented in many types of jobs.

Finally, again as a conservative I recognize an obligation to support programs that work as they are supposed to. And affirmative-action programs actually do work. That's especially true of those established by government contractors under the executive order. But many employers undertake this process on their own; in my consulting work I have observed the operation of such plans for a number of years, and the result has been a substantial increase in the numbers of both minorities and women in jobs that few or none could have obtained earlier. Of course there have been problems with affirmative action, but these tend to be the overdramatized exceptions that make it into TV spots. For the most part, affirmative action has escaped the unintended consequences that plague many social programs, and the progress has come without a great deal of bureaucratic complexity, without imposing unnecessary burdens of time and cost on employers, without creating barriers to employment of white males, and without creating ill will.

In the community of employment managers, testing professionals, attorneys, consultants, and academics who deal with affirmative action, I would expect to find some areas of disagreement with what I have argued, but most would agree that I have made a legitimate case. Why, then, do political leaders and commentators on both sides insist on making affirmative action into an issue like abortion, where the policy options represent fundamental philosophical differences and irreconcilable policy choices? Affirmative action is not a zero-sum game, and its merit does not depend on axioms of political correctness, liberal variety. Maybe we conservatives should actually embrace affirmative action as one of those rare government programs that further our agenda. How about a statement in "The Contract with America, Part II"?

James A. Buford, Jr., is a management consultant, professor, and the author of two collections of essays, The Kindness of Strangers (1995) and The Best of Times (1998), both published by The Blackbelt Press.