Non- Discrimination and Affirmative Action

by Professor Jan Narveson

Non-Discrimination and Affirmative Action

by Professor, Jan Narveson

12th Chapter from his book "Moral Matters" ("Moral Matters" is available from: Broadview Press, Peterborough, Ontario PO Box 1243, K9J 7H5 $20.00)

(reprinted with permission)

These subjects are among the most conspicuous areas for legislation at present, and are closely linked. Both raise serious questions, and are, I think, subject to major objections. Nondiscrimination is the more basic one, since affirmative action is usually invoked as a kind of antidote for what has claimed to be discrimination in the past. Hence we will discuss it first.

We must begin by asking a question about discrimination that is not asked often enough: what is it? There is, by now, a satisfactory short definition: Discrimination is treating some people less favourably than others for morally irrelevant reasons. (The word arbitrary can be used equivalently to irrelevant.) The notion thus defined calls for considerable comment.

To begin with, the phrase 'less favourably' must be taken very literally. You cannot substitute 'worse', for example, without being dangerously misleading. For once we start talking about treating people badly, strictly speaking, then we are no longer talking about discrimination, but instead about simple evil or injustice. For example, we may not torture, beat, or kill people; but the supposed evil of discrimination is not like that. Discrimination in hiring, for example, consists, simply, in not hiring people, though for putatively arbitrary or irrelevant reasons. It involves treating one person less well than another, but not treating the latter badly. To harm someone is to make him or her worse off than if you had done nothing to them. But when you don't hire someone, you are, precisely, doing nothing to them: discrimination in hiring consists in passing people over for a certain kind of exceptionally good treatment - giving them a desired job. Thus it consists in not doing a good, rather than doing a harm. Those you don't hire may have expected you to, and be disappointed when you did not. But that doesn't mean you have actually harmed them, unless they had a right that you not disappoint them in that respect. Yet the normal case is that they have no such right. Giving people a job is not ordinarily something we are required to do; and not giving them a job is absolutely normal. Just now, for instance, I have failed to give you a job. I have also failed to give you two thirds of my current salary, just as I have not done any number of other nice things I conceivably could have done for you. Perhaps you deserve some of those, and the reasons I have failed to give it to you are probably quite irrelevant, morally speaking. Still, as I am sure you will agree, I have done you no harm, no wrong, in all these non-doings. If discrimination is claimed to be wrong, then, it will have to be shown why it is basically different from any of these things. But no one has shown this.

The term 'treat' in our definition brings up the question in what respects considerations of discrimination become applicable. Here again there are major pitfalls to avoid. What if I am literally "treating"? If I walk along the bar, offering a drink at my expense to this person or that one, picking at random from the dozen present, or perhaps picking them on the basis of the color of their ties, am I now to be accused of "discrimination"? Surely not. Again, we see the need for a good explanation of which types of "treatment" could merit the term "discrimination", as opposed to the types that do not. In countries like Canada, the law identifies housing, education, health, and employment as areas in which discrimination is prohibited or limited. Why those and not others, though? Evidently we need to know how such things differ from the example about the bar. Let us take an arguably more important context than drinks at a bar - marriage. Do we have duties of nondiscrimination here? Should the state require us to randomize over potential marital partners, in the interests of "equal opportunity"? After all, almost everyone discriminates uninhibitedly in that context. They marry whom they like, and they like people for remarkably morally irrelevant reasons, such as the color of their eyes or their taste in movies.

Here an important potential snag in our deliberations arises from the fact that in our time and place, there are many services whose provision is required by law. When it is, then the people who provide those services have a legal duty to do so. We have seen, however, that law is not the same thing as morality. We can ask whether the law creating that legal duty in certain cases is doing the right things. Nevertheless, once law has been brought into the act, it is clear that we may not use it for personal advantage, or to play political favourites. If our government is going to provide services, then it has a clear duty to provide them impartially. But that does not prove that morality requires impartiality of us all, in everything we do. It may be a purely political fact, rather than a reflection of a moral requirement, that so many services are now in the "public sector", where impartiality is, one hopes, a sine qua non. Even in the public sector, we should note, the ambit of impartiality is strictly limited. For all of the public services in Canada are offered only to Canadians, and in other countries only to the citizens of those countries. If nondiscrimination were a human right, though, they should be offered without regard to nationality or country of residence. Yet in fact, all governments discriminate sharply against noncitizens, and insist that their citizens do so too. If you are a private Canadian employer, you are not allowed to hire an American or a Hungarian on equal terms with your fellow Canadians, whether or not being Canadian constitutes any kind of qualification for it. In all those innumerable cases in which it does not, how is it relevant to hiring people for my job that the person in question should be Canadian?

The point here, then, is that if there is a duty stemming from the employer's professional obligations, for example as a civil servant, why the service she provides should be rendered impartially, then that is a reason why she should be impartial in that respect. But that has no tendency to show that everybody owes everybody a duty of impartiality. And it also, of course, raises the question whether that public service is one that people really have the right to expect from their governments. We now see that the all-important word in the definition of discrimination is the word irrelevant (or arbitrary). Nobody questions that we may treat some people more favourably than others. The question is what restrictions to place on such differential treatment: which distinctions are such that we treat people wrongly by treating them less well than others in those respects?

We act toward others on the basis of our interests, which are of two kinds: affections - loves and hatreds, likes and dislikes - and other interests - how useful or the reverse they are to us, as in business transactions. These are not sharply distinct, of course. We love children, spouses, friends, and we also derive many benefits from our interactions with them. Clearly, our affections will prompt us to treat people we like better than ones we don't; similarly, our other interests will prompt us to extend more opportunities and benefits to those who benefit us in turn. One would expect, then, that a person's bestowal of favours would be roughly proportionate to his degree of favourable inclination and the amount they benefit him.

People who think that there is something inherently unjust about discrimination might think that we shouldn't be dealing with people on a basis of self-interest at all. Some religions, and most notably the Christian, have been thought to imply some such thing. That is certainly a misperception in the case of Christianity, for that religion teaches people that if they love others, God, at least, will love them even if no one else does, and will and reward them in Paradise if not before. That is plainly an appeal to long-range self-interest. Nor, I think, will examination of other religions reveal anything markedly different. How could it? Total, uncompensated self-abnegation is absurd, and we should be extremely suspicious of the motives of those who claim to teach it; characteristically, our unselfishness turns out to redound to their advantage. Be that as it may, the point of view of this book is that self-interest is wholly legitimate, that curbs on our pursuit of our interests are curbs on short-sighted, narrow, ill-judged pursuit of those interests, but nevertheless based on our real, long-run interests. Morality, in particular, bids us take account of others, who think much as we do and are similarly self-interested, and to adjust our actions accordingly.

At some ultimate level, perhaps, fundamentally arbitrary distinctions in our treatment of others may be irrational. If there is literally no difference between B and C, how could I treat B and C differently? But even that has a good answer in many case: I might deal with B rather than C because I simply can't deal with both, I must choose, yet have no reason to prefer one to the other. So I just choose, by chance or whim. Walking along the street in Cairo or New York, one might impulsively give a dollar to this beggar rather than that one, for no other reason than that one happens to have one dollar and not two, and doesn't think it worth bothering to make change so as to give them each 50=A2. But isn't that perfectly reasonable? Of course it is, and one does this constantly. The man you end up marrying, if any, will probably be one whom there is no more and no less reason to choose than some other; but you choose him because you want to marry, and he happens to qualify and be available. That is a perfectly reasonable thing to do. By contrast, it would be unreasonable to insist that you marry only the Perfect Man - by the time he comes into view, you will have been long in the grave, and endured a lonely existence in the meantime.

On the other hand, you of course choose the best man from among those available. If you think that Jones would make a better husband than Smith, you of course will marry Jones, in the absence of very special reasons to the contrary. The same goes for employing someone. Choosing genuinely arbitrarily, with absolutely no reason to do so, is virtually unintelligible. So if the point of denouncing "discrimination" were to be sure that we don't engage in such activity, then we would be wasting our time. People always have some reasons for what they do, even if only that they must do something and lack time, inclination, or resources to make fine distinctions. And if they act for ill-considered reasons, then they act unwisely - but no charge of injustice applies just on that account; we cannot be free without being free to make mistakes. There is, then, no foundation for principles of nondiscrimination in the structure of practical reason, just as such.

Comparative Advantage

However, we do need to pay attention to an interesting and treacherous set of motives: jealousy and envy. People are often sensitive not only to how well they themselves are doing, in various respects, but also to how they stack up against others. The man who sets out to be the "fastest human alive" is not interested merely in how fast he himself runs, but in whether he is faster than others. If we are dealing with a number of people who are motivated to a significant degree by these comparative interests, then if we do not treat one person "the same" as that person's associates in the respects they are interested in, we are likely to be in for trouble. There will be complaints, bad tempers, whisperings, and the like.

Suppose you have a business and must decide how much you should pay your various employees. What principle will you use? Well, first, you will surely want to pay your good workers enough to keep them from moving somewhere else. It would be silly not to, for if they leave, you will have to find new employees who may be less capable, and in any case it may be quite awhile before they can perform as well as the veteran you have let go. Secondly, you will want to pay more to those who are worth more, as measured by their contribution to the success of your business. Obviously this interacts with the first factor, for if you offer less to someone who does more, then he or she will be motivated to go elsewhere; but if you overpay someone, you are losing money - you could have hired some equally productive person for less money, and pocketed the difference. Since the point of the enterprise is to make money, such practices are self-defeating.

So far, so good. But suppose that some employees insist on getting just as much as others, even though, in your judgment, they don't do as much for you. Do you pay them more, thus making your operation less efficient? Or do you pay them less, thus risking their disaffection? The latter might lead to poor performance, even sabatoge, on their part; the former might lead to ruin. So you have a problem. There is no simple solution to it. But there is a relatively simple abstract principle to invoke: you solve the problem by finding the compromise between the two contrary forces that will make your firm maximally profitable over the longer run. You will likely pay some people a little more than you really think they're worth, for the sake of keeping harmonious relations among your employees and thus sustaining morale, and consequently output.

Now, there may also be some among your employees whom you would like to pay more because you simply like them more, and not because they are worth more. This, too, will cost you profits - but perhaps you are willing to spend the money anyway. Is this something to be prohibited? If it is, could anything but envy or jealousy be the reason for the prohibition? The question about discrimination comes into sharper focus when we consider such examples.

The Puzzle about Nondiscrimination

We don't owe people jobs - they aren't among the things we have a basic duty to provide for others. Those who are in a position to distribute jobs are, mostly, in business, and their companies create the jobs in question; jobs don't grow on trees, for the picking. Those companies were established in order to make money, by selling goods or services. Nobody may require anyone to do that: whether we want to go into any sort of business or offer anyone a valuable service is up to us, not anyone else.

But if it is up to us, then how are we are supposed to have a duty to offer jobs in our business to A only on equal terms with B? Again, consider our earier comparison with marriage. Do we have a duty to treat candidates for spousehood "equally"? Am I supposed to give Agatha and Belinda equal chances, an "equal opportunity", for my hand in marriage? Surely not. Similarly, if I am inviting people over for dinner, or getting a bunch of people together to play ball, or to help me build my summer cottage, I plainly have no moral duty to invite the general public, randomizing over friends and total strangers alike.

Well, if I had no duty to go into business, and it's my business, then why must I randomize over women and men, whites and blacks, or whatever, when it comes to handing out jobs? Why can't I hire whomever I please? It is useless to respond by saying, "because you have a duty to treat people equally", for whether that is so is exactly what's in question here. Moreover, the examples of hands in marriage, handouts to arbitrarily chosen beggars, and any number of others one could readily produce, strongly suggest that it is false. We do not have to treat people equally. We may and should treat people we like and people who do more for us, better than others.

Perhaps it will be said that the fact that person B is my friend is what makes it relevant to treat him better than some unknown. But why is it relevant? Those many people who think we have a duty not to discriminate against others think that the fact that B is a friend of mine is not a good reason why I should be able to hire B instead of C: that, they insist, is "irrelevant". And in fact, what is called discrimination, generally speaking, consists in preferring some groups of people over other groups just because you like them better. But why should we think that wrong? What's wrong with it?

The fact that the ones you have preferred will end up slightly better off than those you haven't is perfectly true - indeed, that was the point of your treating them better. But then, did the unpreferred ones have any right over you, to end up equally well off from your choices? There is no obvious reason to suppose that there is any such right. Indeed, there seem very clear and compelling reasons to suppose that they there is not.

Competence

Most people who talk about discrimination in hiring think that what is relevant to hiring is competence, the ability to do the job. And indeed, they go on to insist that people from the groups allegedly "discriminated against" are just as good at the task in question as people in the group allegedly favoured: black people just as competent as whites, and women as men, for instance. There is an oddity, indeed a basic paradox, about this claim, taken as a reason why we have a duty to refrain from discrimination. Duties normally curtail action from motives of self-interest. If you have a business, then it is in your interest to hire competent people, or more precisely, to hire the most competent person you can get at a given wage. If B is better than C at the job, and you can get B for the same wages as C, then it's in your interest to hire B, and as a good business person, that's exactly what you will do. For if you don't - if, for instance, you insist on hiring your incompetent friends instead - then your costs of doing business will be higher than they need be. If you do enough of that, you'll be uncompetitive against your more rational rival businesspeople who don't care about the race, sex, national origin, religion, of other characteristics of the people they hire, so long as they get the best worker at the lowest cost.

Discrimination is therefore inefficient. It costs money. If your sole motive is to make money, you will not discriminate. Yet letters to the editor, political speeches, and any number of laws passed by legislatures are designed to get people, and in the latter case to force people, not to discriminate. But why would all the pressure, and the use of force, be necessary if the aim was to get them to do what makes them more money -- that is, precisely what they want to do anyway? It doesn't make sense! On the other hand, suppose that their aim wasn't to make money? Suppose the organization you are hiring for is devoted to something else, such as the welfare of some particular group - your school, say, or a church or club? Would we still have to hire the "most competent", even if those competent people weren't interested in the purposes of the club or the church in question? Does the Presbyterian Church down the street have to consider Catholic and Presbyterian applicants for the position of janitor on equal terms? Why? What the manager of a club, church, or other nonbusiness group wants is to promote the purposes of the club, and from that point of view also, the enforcement of so-called "nondiscrimination" is arbitrary and wrong. It is interesting, and ironic, that the very thing discrimination was supposed to consist essentially in -- namely arbitrary treatment -- is precisely what an antidiscrimination principle seems to entail.

To tell an interest group that they must ignore their interest when it comes to hiring would be to do what governments have no business in doing: namely, to tell people how to live their lives, rather than letting them live the life that they choose. If the interest is in making money, then to tell a business whom to hire and whom not, is either to tell it to do what is most efficient and hence makes it the most money, which is absurd because redundant, or it is to force it to engage in inefficient practices, thus denying it the right to do what it exists to do. Either way, the intervention is wrong. People do not have a duty to make money; we have no business forcing people to maximize their incomes or profits. And on the other hand, so long as the business or other activity they are engaged in is inherently legitimate, providing people with goods and services that do not, of themselves, involve the doing of evil to anyone, then we also have no right to curtail their pursuit of this in the interests of some other cause.

In recent years, nondiscrimination has changed course. Instead of merely insisting that people not discriminate on the basis of supposedly irrelevant characteristics, enthusiasts for equality now insist that workers in employment X must be half male and half female, and that the fraction of those employees who are of one race or ethnic group, or whatever is the currently fashionable alleged discriminandum, be equal to that group's fraction in the whole surrounding population. But plainly, this is entirely arbitrary from the point of view of those who create and sustain those businesses. What the businessperson wants is to hire the people who will make the most money for the business, be those people male or female, black or white, or whatever. If those who can contribute the most should happen to be of one color, sex, or ethnic background rather than another, then employees in that business will tend to be predominantly of that kind. Here again, to require that people pay heed to such structural characteristics of their work force is arbitrary and invasive.

Forced Affirmative Action

There is a problem, then, about giving any very good reasons why what is called "discrimination" should be considered evil or immoral at all. But now let's suppose, for the sake of argument, thatere is such a duty to treat people "equally". Let's also assume that in the past some people have not been "treated equally" in those respects. Should we now try to help matters out by actively promoting the prospects of the formerly discriminated-against group, by requiring employers to employ some larger proportion of that group than they would otherwise do, or have done in the past?

Notice that the question before us is not whether any such programs, by anyone, are ever allowable. In fact, what makes it obvious that they can be is the very argument we have just been constructing regarding discrimination. Affirmative action discriminates; but if discrimination is morally permissible anyway, then that would not be a fatal objection to it. However, the question is not whether it's inherently all right to practice it. Rather, our question is whether it's all right to impose such a program, by the coercive powers of government. The question, in other words, is whether justice ever requires such a program.

A Fallacious Argument for Affirmative Action

Those who think that affirmative action is required by justice often use fallacious arguments. One especially popular one is that affirmative action will help to correct the past injustices that allegedly occasion it. But that is not so. The only way to do that would be to find the particular people who lost whatever they didn't get by virtue of discrimination and then see to it that they get it after all, or some suitable compensation. A lot of those victims are dead and beyond compensation, while in many other cases, other difficulties would stand in the way. In any case, the problem with expecting a program of "affirmative action" to help out is that it doesn't do that. Instead it rewards new people, people who in most cases have never been discriminated against. And it rewards them excessively. For if the problem was that they weren't treated equally, then the right thing to do by way of "compensation" would simply be to treat them equally. But affirmative action treats them unequally. Instead of ensuring that they get the same chance as others, it gives them a greater chance than those others. If in the past men got the jobs instead of deserving women, clearly the cure would be to open the doors so that the deserving women now may get them as well - rather than shutting the door on the men, deserving or not, thereby giving women greater chances than men for such employmnt .

Moreover, a program of forced affirmative action penalizes employers who have never discriminated against anybody, by forcing them to consider more favourably people they might not otherwise have hired. This is often obscured by claiming that the program only requires employers to hire the "equally best qualified", rather than to lower their standard. But what if the nondiscriminating employer hires more women than men? Suppose that he has done so because more of the best candidates were women? But no program of affirmative action is going to rest content with that. It's going to insist that the employer is "subtly" discriminating after all. It may, and probably will, impose a quota, regardless of any evidence about the proportions of males and females among the most-competent individuals. If the reason for the program was a belief that competence is equally distributed between the group favoured by the affirmative action program and the other groups not so favoured, then the basis of that belief is no longer the experience of employers but an a priori belief by the government that it is so distributed. It will end up forcing employers to take people whom they didn't believe were as desirable as the ones they would otherwise have hired if they'd had their choice. And there will be no market test imposed on the government to see whether their wisdom on this matter has proven correct. That test will be imposed exclusively on the employer, and since it is the test he faces daily in business, the imposition of an a priori criterion of this kind is highly likely to be wrong from the start. The businessman will be forced to pay the cost. Or his customers. And, in the end, his other employees, who will share in diminished income or bear the brunt of plant closures and contractions if the result is disastrous.

Affirmative action programs also require administration. If we're going to force people in businesses X, Y, and Z to be egalitarian in their hiring, then overseers will have to be appointed to see to it that they do so. The people sent out for that purpose will not, of course, know as much about the business as its managers and owners - after all, they've never been in that business. Nevertheless they will presume to instruct the actual managers and owners, deciding who is and who is not "equally qualified"! Inevitably, this will have adverse effects on efficiency. The numerous administrators required to superintend the program have to be paid by somebody, and the less-efficient employees hired because the employer was forced to hire them will produce less. Both of these factors conspire to reduce the amount of goods and services available for people who would have liked to have them, and to raise the price of the ones that are produced.

Equal Pay for Equal Work

A similar idea to affirmative action is the enforcing of a principle of "equal pay for equal work". There is, again, an oddity about the whole idea, for paying equal amounts to people who are equally profitable is precisely what the market situation of a business induces it to do anyway. It already is paying "equal pay for equal work", as far as it is concerned. But the business firm uses the right criterion, marginal effect on profitability, rather than the irrelevant criterion imposed by the rule: hiring, for example, on the basis of who expends the most calories rather than on the basis of who makes the greatest overall contribution to output per unit of wages. But the administrator, or the lobbyist, or the member of the group that is supporting the allegedly discriminated-against people, doesn't think that marginal profitability is the appropriate criterion. He or she will come along with some new ideas about who is better than whom at that job - ideas engendered in the political climate of the time, not by any rational assessment of the needs of the business. This, despite the fact that they by hypothes know much less about the job than the people who now own and manage the firm.

So again, what will happen is that the business will become less efficient. It will either have to pay the costs of the administrators, in which case its customers will pay, or else the taxpayer will. But if the taxpayer does, then that will leave him with less money to buy the products of this company, or of any other. The effect is the same: higher prices, and thus a lower standard of living, for the people served by those businesses. In addition, the people who have now lost out on jobs they otherwise would have gotten will hardly be very happy with the new state of affairs. They will resent the beneficiaries. And meanwhile, the people now hired, who are perceived as having been hired for reasons of affirmative action rather than because they are the most able people available, are likely to be looked upon with suspicion or condescension by their new colleagues. In general, everyone loses, except perhaps the special group that it is the aim of the legislation to benefit - for awhile.

footnote: "See the sobering study by Thomas Sowell, Preferential Policies (New York: William Morrow, Quill book, 1990, especially pp. 76-87 concerning the war in Sri Lanka."

Why should the government have the right to do any such thing? Why should it benefit one group in the population at the expense of the rest? These are not trivial or merely "academic" questions. Wars have been caused by just such measures. When you draw a line dividing a group into two, and force people to treat one of those groups more favourably than the other, the losers become resentful, and the winners self-righteous and protective. Both resort to politics and coercion instead of peaceful production. None of this makes for smooth human relations.

Determining The Distribution of Competences

To make any assessment of discrimination on a group basis, one has to know two things: first, the proportion of those in the allegedly discriminated-against group who were in fact hired (or whatever the relevant mode of treatment is - hiring is our subject here), and second, the proportion who should have been hired. How are we to know the latter?

Sometimes one can have "anecdotal evidence": first-hand observation by someone of a case in which an allegedly less competent individual was preferred over another individual who was presumably superior. Anecdotal evidence has two major disadvantages. First, it is necessarily far too limited to afford any secure generalizations about the large classes affected by any affirmative action program. Second, those who tell the anecdotes are very likely to have their own axes to grind. How does the teller of the anecdote "know" that the successful applicant wasn't better than the rejected one? And if it is disputable, why should we believe the relator of the anecdote rather than the employer, who, after all, hias an interest in hiring the person who will produce the most profit?

Those who support coercively imposed affirmative action, while of course influenced by anecdotal evidence, are primarily moved by their perception of structural disparities. They note, for example, that more men than women are hired in a certain line of work. Since the numbers of men and women in the population are about equal, they infer that something is wrong. But the inference doesn't follow directly. From "The population is x% F's" and "The percentage of persons employed in work W is less than x", we cannot infer that F's were discriminated against unless we have reason to believe that the percentage of F's in the class of most-desirable candidates for W is the same as it is in the general population. And there is often excellent reason for expecting that not to be true. Indeed, in the special case of men and women, there is one extremely important reason for expecting it not to be true. For we are not counting marriage as a case of "hiring" - even though there is really very good reason for so counting it, in many respects. But if we don't, then all of those women who marry and have families instead of going to "work" - that is, joining the paid labor force - reduce the size of the female hiring pool available for all other occupations. So long as more women do that then men, the result must be that we cannot assume that any other given occupation should show an equal distribution by sex among its employees, absent which we can impute discrimination.

Nor is that all. What if the F's are typically not very interested in making their living at that particular job? Suppose that typical women just don't like to work in coal mines, while typical men just aren't much interested in secretarial work? Doesn't that matter? It need have nothing to do with anything as abstract as "native ability". Whatever the reason for these differences in interest, they will affect the structure of the available pool for that kind of work. And whenever this is so, it will be wrong to force employers, in the supposed interests of "equality", to hire a similar percentage of males and female for that employment. Of course, the same holds for any other groups as well: groups of different ethnic background, race, or any other arbitrarily selected characteristic can be expected to have different patterns of preference, as well as different specific skills.

There are innumerable differences betweeen groups besides the defining differences, and there is no reason why some of those should not be relevant to occupational choice and suitability. The number of black males on American professional basketball and football teams is vastly greater than the proportion of blacks in the general populace, while the number of Orientals is vastly smaller. Is this due to discrimination? Certainly not. The difference in height and other physical attributes alone would account for it. In hockey, the great majority of professional players on American teams are Canadians, to the surprise of no-one who knows anything about the climates in the two countries. On the other hand, the number of women employed in the hand-assembly branches of the electronics industry, or in sewing, is vastly greater than their 50% incidence in the general populace. Is this due to discrimination? Again, the sheer physical differences of the two groups is enough to account for it: the smaller and more delicate hands of typical women are obviously better suited to manipulating tiny physical objects. Examples could be multiplied indefinitely. This problem is so enormous as to be, in all likelihood, insoluble - that is, insoluble in any way that is favourable to the argument for affirmative action programs. That is why those who argue for them just assume, without argument, that certain proportions "obviously" ought to be equal, and let it go at that. This a priori attitude is hardly appropriate to real life, where everything depends on the facts. It is absurd to make such claims in this field; philosophers have long been aware that the human intellect is simply not the tool to rely on, exclusively, when what is at issue is how things are out there in the world.

Achieving Other Goals by Affirmative Action

One major source of support for affirmative action, very likely, consists in employing it as a device for achieving some other social goals. Someone may support the imposition of a quota of 50% women in a certain area because he thinks that in the ideal society, that's how that area would be handled. Never mind that there are not nearly enough women interested in or available for that kind of job as things stand to staff 50% of the available positions, or that there is no reason to expect that there ever will be, so far as our best current information is concerned: our theorist has the Ideal Society all figured out, and he's jolly well going to make us all go along with it if he can. But we shouldn't allow that kind of blackmail or Utopian planning. We must deal with people as they are, and not with people as we or any particular group think they ought to be. Realistically, it makes all the difference that the positions in question are not popular with women. It's not our business to revise human nature to suit our theories.

Conclusion

It is, all in all, very difficult to justify either forced anti-discrimination or forced affirmative action in the sense in which they are now popular. Our leading principle in this book has been that people in general should, in the absence of good reason, have the right to do as they want, work at or produce what they wish, and associate with their fellows on terms of agreement rather than by force. Those who don't accept this general account of morals should be asked on what they propose to base their visions, and why they should be allowed to impose those visions on ordinary people trying to make the best lives they can for themselves. We should not be impressed with any of the arguments typically employed in those directions, all of which seem to be either riddled with outright fallacies, vague at crucial points, or just arbitrary, high-flown pronouncements. The rule of liberty stands, in business contexts as in the others we have explored.