|Wednesday, March 18, 1992|
Alternatives to job quotas
Well-defined jobs specs will reduce bias, get best employees
© Charles Maurer
Last of three parts
The Government of Ontario is proposing to pass legislation on employment equity. This means primarily defining quotas for hiring and promoting aboriginal peoples, persons with disabilities, racial minorities and women.
The last two days, I showed that quotas of this sort are iniquitous, that similar legislation has not worked in the United States, and that the government's proposals in particular are conceptually incoherent and unworkable.
Today I want to propose an alternative approach.
To begin, let's pickle some red herrings. Discrimination, per se, is not a bad thing. It is vitally necessary. If you need your appendix removed, you bloody well want to discriminate between a surgeon and a plumber. Indeed, you want to discriminate between an abdominal surgeon and an orthopedic surgeon, and if possible, between a better abdominal surgeon and a lesser one.
These discriminations are appropriate and fair, both to you and to the people you are discriminating between.
But you would be a fool to discriminate between surgeons by ranking them in alphabetical order. You might harm yourself by favoring a hamhanded Dr. Alpha instead of a facile Dr. Zed. Moreover, this would be unfair to Dr. Zed, for her father's name does not affect her skills nor could she choose her father.
On the other hand, let's say you want to hire a file clerk. Two people apply for the job, an unemployed English professor and someone whom schools used to call 'retarded educable.' You would be foolish to hire the professor for he would be bored to distraction. He would become the office sourpuss and in his boredom and pique would make mistakes. But a retarded fellow who knows his alphabet might find the job challenging and enjoyable and do it happily and well. True, the professor has higher qualifications so rejecting him would be unfair. Yet hiring him would be unfair to the business.
In short, discrimination in hiring is desirable, not wrong. Unfair discrimination is certainly not good but when people have conflicting interests, fairness to all is impossible to achieve. The important factor in hiring -- the factor that ought to be sought by law -- is not non-discrimination or fair discrimination but appropriate discrimination.
To foster appropriate discrimination, I propose a mechanism that would define sensible specifications for individual jobs. Once that mechanism is in place, I propose that the law should require firms to hire and promote people according to these specifications.
Nothing is easier than to define objective criteria for a job. People do it all the time: 'A file clerk needs a high school diploma and must be able to alphabetize 100 names in 10 minutes with no mistakes.'
The trouble is, although nothing is easier to define, nothing is more difficult to validate. You can define criteria, as personnel officers do customarily, by fiat; you have to validate them by tests. And tests are not easy. Designing them is trickier than brain surgery.
People are trained to do this kind of thing in the PhD programs of university psychology departments. Few PhDs now specialize in personnel psychology because there are few jobs in it. But I propose that this be changed.
I propose the government do two things: (1) Create a licensed specialty at the doctoral level in personnel psychology. (2) And require that firms larger than some minimum size engage licensed personnel psychologists to define objective specifications for all jobs as they become available.
This would be slow going at first. Few psychologists would be available and they would have to spend a lot of time developing tests. This means that initially the policy would be followed only by the government. But the number of psychologists would increase rapidly and with them a body of proven tests.
The result, ultimately, would be the equivalent in employment of what we now have in construction. The structure of every building over a certain relatively small size must be specified by a professional engineer. The engineer takes legal liability for the suitability of his specifications. During construction, the government sends out building inspectors to make sure the specifications are met.
Although requiring licensed psychologists to 'spec' jobs might sound unwieldy, in practice it would be no more awkward than requiring professional engineers to spec a building. Someone has to define a job's specs anyway. I would just require that the specs be put down on paper and be written by someone who has been trained in defining them appropriately.
Nor would this entail added expense. Psychologists would keep on hand specs for common jobs so these would be cheap to obtain. More unusual jobs would be more costly to define but they are now, too: you would merely exchange an untrained personnel consultant for a trained one. Businesses would recoup the marginal extra costs by filling their jobs with more appropriate people.
Indeed, many businesses would benefit because they would do away with the inappropriate ways that personnel decisions are commonly made now. Within a bureaucracy, someone who does a good job is likely to be promoted, and to be promoted again, until he reaches a job that is too difficult for him to do well. Once he reaches that level, he is no longer promoted but neither is he demoted. He stays put forever. In this way the managerial levels of a bureaucracy become filled with mediocrities. (This is the Peter Principle.)
From these mediocrities more mediocrity spreads. One thing a mediocre manager does not want is a hot-shot beneath him in the bureaucracy. So mediocre managers tend to hire and promote mediocrities.
If jobs had appropriate specs, none of this would be as likely to happen. It would be easier for a manager to see that, although Jo is a fine sales manager, she hasn't the numeracy to be a good vice-president -- or that Bill, whom Louis wants to make his new assistant production manager, does not really know enough about computerized assembly lines. With this information, businesses would become more efficient.
If a building's structural specifications are faulty, the engineer is legally responsible. If a faulty design lets the building collapse onto its occupants, he will have to pay for the repairs and he might even be jailed for criminal negligence.
If the specs are implemented wrongly -- if the building collapses because someone didn't bother to attach a beam -- then the builder must pay to have it fixed. Moreover, the carpenter who decided that the bolts were superfluous might be jailed for manslaughter, and the foreman who should have noticed can be jailed for criminal negligence.
I would hold people responsible for personnel decisions in exactly this way. The personnel psychologist would be responsible for defining appropriate specifications according to the norms of the profession. Businesses would be responsible for implementing them.
If a company ignored specs, it would be liable for civil damages much as it is now under anti-discrimination legislation. Moreover, if a prosecutor could show to criminal standards of proof that ignoring specs caused a substantial personal harm, then a personnel manager could be fined personally and so could the vice-president who should have noticed and prevented the illegality. Repeated offences would send the manager and the vice-president to jail.
To monitor hiring, employers should be required to maintain records for some number of years showing each job's specifications and how each applicant fit them. From time to time, personnel auditors would come by -- as building inspectors do -- to check them.
Despite all this, Acme Widgets might still try to avoid employing women whenever it can, to avoid interruptions from maternity -- especially if it must pay for maternity leave.
Ditto with paraplegics and the palsied. No matter what the law says, Acme will somehow find a way not to hire them if accommodating them will cost money.
Yet society needs women to bear children and it needs the paraplegics and the palsied to be employed.
Individual businesses should not have to pay for these societal needs and business people will always wriggle through loopholes to avoid it. It would be better that society pay as a whole, through taxes.
Taxes, not individual businesses, should pay for maternity leave, and taxes should help to pay for creches and day care. Taxes should also pay for special accommodations at the workplace that handicapped people might need.
This approach to employment equity is not perfect, of course. Licensed experts are still people, people who reflect many of the prejudices of society at large. Moreover, they will have to define the nearly undefinable.
But in at least the more scientific branches of psychology, training at the PhD level is training at recognizing nonsense and bias in disguise, and in dealing empirically with the undefinable. Personnel psychologists trained rigorously at that level would come to be able to do a reasonable job.
Certainly my approach stands a better chance of working than quotas. If nothing else, it addresses the correct problem, ensuring that discrimination is appropriate. And it addresses the problem using a model that works in another field (construction).
It would also be less costly to implement and easier to enforce. Engaging specialists to spec jobs would add little to the cost of hiring and promotion. Many businesses would actually save money overall by filling positions with more suitable personnel. This would be a carrot for businesses to co-operate.
In contrast, corporate machinery to define, implement and report on quotas has proven to be complex, costly and, from the businesses' point of view, a waste of money. The government needs constantly to wield an impracticably large and costly stick.
Moreover, litigation would be easier with my approach -- easier and hence cheaper. Statistically undefinable quotas are always arguable but specs would be there in black and white. So would applicants' ratings.
My approach would take time to implement, of course. The maternity leave, day care and aid for people with disabilities we could fund with taxes now; but most of the plan would take a generation to effect fully. This would be too slow for many.
But if it would take that long, at least it would have some chance of working.
In contrast, steering society with quotas is like steering a ship by pounding with a sledgehammer at the helm. You will make a big noise and break a lot of fittings but the ship will not change course. All you end up with in the end is a mess and a bill for repairs.
If we implement quotas, a generation later we will be in exactly the situation that the United States is in now: billions down the tubes to try to enforce them with nothing to show for it but outrage from whites and support for the likes of David Duke.