The Case for Discrimination

23. DIRECTIONS FOR FUTURE RESEARCH IN EQUAL PAY LEGISLATION

THERE ARE MANY CANADIANS WHO NOW SUPPORT EQUAL-PAY legislation.1 As one indication of the popularity of this idea, the three major political parties in this nation seem to have adopted for themselves several of the major planks of this program.2

But there are problems. If legislation incorporating equal pay for equal work and equal pay for work of equal value is enacted, the present (mainly) marketplace determination of wages will inevitably tend to be replaced by the arbitrary edicts of civil servants, bureaucrats, consultants, judges, and/or human rights boards.3 In contrast, one of the important functions of wages in the market system is to allocate labor to its most needed and productive locations. If this process is short-circuited by equal-pay-for-equal-work (EPFEW) and equal-pay-for-work-of-equal-value (EPFWEV) legislation, then labor mobility to that extent will be reduced.4 But it is important that workers locate themselves in accord with changing consumer demands. Unless they can be induced to act in this manner by market signals in the form of wage changes, the flexibility of the economy will be diminished.5 As well, equal-pay enactments function so as to “protect male jobs from low-wage female competition.”6

Things would be bad enough if EPFEW and EPFWEV were required to right wrongs now existing in the labor market. Then, we would have to face a tradeoff between the injustice of discriminatory behavior imposed on women by employers and the risks of economic inefficiency. But a drawback of equal-pay legislation is that women are not being victimized in the labor market by the discriminatory practices of employers. Thus, such legislation will not be effective in its main object: it is a cure for which there is no disease.

According to the most recent statistics available, the ratio of female to male earnings in Canada is 64.7 At first glance, this might appear as prima facie evidence of the existence of employer discrimination against women. But, while superficially plausible, such an explanation is highly untenable. In order to see this, let us assume that male and female productivity in the marketplace is exactly equal. If so, successful discriminatory employer behavior would entail that women, not men, were being paid far less than their marginal productivity levels. Should this state of affairs ever occur in the first place, it would be very unstable. For large profit opportunities could be gained by those willing to employ women. Sex-neutral entrepreneurs could drive to the wall those who insisted upon indulging in their “male chauvinistic” tastes for discrimination.8 And in the process of “exploiting” the poorly paid women (by hiring them and bidding up their wages), these profit-oriented businessmen would act so as to equate male and female incomes.9

But the facts are clear. Women do earn far less than men. If this is not due to employer discrimination,10 how then can we account for it? There are several alternative explanations. The first and most basic is that the supposed male/female income gap is really nothing more than a statistical artifact. Consider married women first. One problem is that a “family’s income is recorded in the official statistics under the husband’s name”11 alone. But most marriages, at least in their economic aspects, are like a business partnership. The husband may earn all or most of the income in a superficial legalistic or accounting sense, but it is due, in great part, to the wife’s efforts that his salary is as high as it is. It is therefore highly misleading to credit the husband with all or most of his “own” income. It would be much more accurate to divide total family earnings by two and credit each marriage partner with a full-half share.

Suppose there were two attorneys in a partnership who agreed to split the proceeds of the firm equally. A, the “outside” partner, deals with clients, conducts the trials, and brings in new business. B, the “inside” partner, looks up the precedents, does the research, and manages the office. To credit A with all or most of the profits, a ludicrous supposition, makes exactly as much sense as assuming, as does Statistics Canada, that the typical Canadian husband really earns all or most of his “own” income.

In point of fact, husband and wife act in many ways to enhance the registered income of the former and reduce the registered income of the latter. This is done in order to maximize family earnings, given other family desires, such as raising children. Examples of such behavior are numerous. While this may be changing slowly, at present a married couple will typically choose a geographical location to enhance his earnings, despite what it does to hers.12 As well, there is almost always an unequal division of child care and housework responsibilities.13 There are differing labor force participation rates,14 education and training,15 and advanced degrees.16 One indication of the strong asymmetrical effects of marriage on registered earnings17 (increasing the male’s, reducing the female’s)18 is that the female/male earnings ratio for those who have never been married is a startling 992.19

A third explanation for the female/male wage gap is occupational choice: women tend to enter lower-paying occupations than men tend to enter.20 Here, human capital obsolescence—because of time off for childrearing—will not occur to as great a degree21 and will not penalize part-time work as stiffly.22 Occupational choices toward low-income careers (“pink-collar ghettos”) are also made by unmarried women. Partially, this may reflect anticipated married status in the future. According to some analysts,23 this choice may be due to women’s lower self-esteem, or self-image, or fear of success. In their view, young girls are socialized into believing that they are inferior to boys, and that they must at all costs avoid competing with males. If true, this phenomenon could account for lower expectations and ambitions. On this basis, wives might reject raises or promotions and avoid entering higher-paying occupations in the first place,24 for fear of making themselves unattractive to their husbands, present or future. To the degree that such behavior occurs, it is a personal tragedy for the women involved—psychologically, socially, and personally—in terms of the human potential destroyed. But the explanation for this must be complex and deep-seated: it can hardly be blamed on employer discrimination in the labor market.

There is a reason why the phenomenon of unequal wages between males and females seems to be in need of explanation. It is because of a basic assumption that, in the absence of discrimination, male and female earnings would be equal. And underlying this is the view that men and women have equal productivity in the labor market. (With unequal economic productivity, unequal wages would not be in need of any explanation.) But this is more of a pious hope than it is a conclusion based on evidence. That it should be taken as an article of faith that male and female productivities must always and ever be equal has more to do with political ideology than with the realm of economic reality.

So deeply entrenched is this view that it even spills over into methodology. In many econometric and empirical works, any male/female income differential that cannot be accounted for on the basis of variables such as age, education, labor force participation, and so on, is assumed to be the result of employer discrimination. Discrimination, that is, is seen to be a “residual”: if gender differentials cannot be explained any other way, they are accounted for on an a priori basis by discrimination.

But there are grave problems with such a view. First, “it would seem evident that the failure to explain the wage gap by a given set of variables is consistent with the operation of undiscovered variables having nothing to do with discrimination.”25 Second, the “human capital” variables employed in most regression analyses26 of this type are only highly imperfect approximations of what really accounts for productivity. Years of schooling, for example, admit of great differences in quality. Their correlation with productivity is far less than exact. Third, this imparts a bias toward that which can be quantified as an explanation of the gender pay gap. Ruled out of court as unquantifiable are such things as ambition, perseverance, motivation, pride in being a breadwinner, reliability, competitiveness, attitude towards risk,27 and, dare we suggest it, possible innate, biological, sex-linked differences.

Let us now return to equal pay for equal work and equal pay for work of equal value. We have already seen that one drawback of such enactments, as with all legislation interfering with the market process, is the tendency to retard the ability of the economic system to allocate labor to its (continually changing) most optimal employments. In the case of equal pay for work of equal value, third-party “experts” will be called upon to determine whether mainly male occupations, such as that of truck driver, are “really” of equal value to jobs held mostly by females, such as secretary. A spurious scientific objectivity will be imparted by numerically rating such aspects of these callings as training, responsibility, working conditions, education, and so on, and then adding them together to derive a total point score. Say what you will about such a scheme, at least it has one undoubted advantage; it will serve as a full-employment measure for lawyers, for the values assigned to each dimension can only be arbitrary. The procedure will thus open up society to a spate of contentious lawsuits, as the various newly created pressure groups endlessly strive for more favorable ratings.

The point is that there is no such thing as an intrinsic or objective “worth”28 of a job (nor of goods and services such as paper clips, music lessons, and so on). On the market, crucial in the evaluation of employment slots are the subjective rank orderings of the consumers—the willingness of people to pay for things. The job of whip-maker, horse-trainer, or carriage-wright might have required tremendous investments in skill and great responsibility. But with the invention of the horseless carriage and fickle consumer preferences, all this goes for naught. Were there such, the expert job evaluators at the turn of the century might have given these tasks high point totals. But on the market—that is, in reality—these jobs were suddenly rendered obsolete and valueless.

Presently, the jobs of dentists, dental hygienists, and teeth x-ray technicians all require much intelligence, years of intensive training, great diagnostic skills, and a high level of professionalism. Were the evaluators unleashed upon these jobs to work their magic, there is no doubt at all that a high point total would ensue. But if and when a cure for tooth decay is found, these skills will go the way of the dodo bird, as far as value is concerned. Consumers will no longer be willing to purchase their services, and the returns to human capital invested in these lines will fall precipitously.

Let us consider one more example. Suppose that female prison guards do exactly the same quality and kind of work as that done by male prison guards. We assume, in other words, that male and female prison guards do “equal work.” But let us suppose that, for some reason, women are far more reluctant to enter this profession than are men.29 Under such conditions, in the marketplace, female prison guards will receive higher salaries than their male colleagues. This, according to the logic of the equal-pay-for-equal-work philosophy, is obviously “unfair.”

What can be done? If the female wage rate is lowered to that of the male, there will not be an adequate supply of women prison guards to satisfy the demand. If the male wage rate is increased to match that of the female,30 there will be an oversupply of male prison guards. If the wage rate of both is set at some intermediate point, there will be an excess supply of men prison guards and a shortage of women prison guards.

If the expert evaluators also take into account this phenomenon in their evaluations of male and female prison guard jobs (as well as all other unquantifiable factors that determine wage rates), they will escape the quandary of creating either a shortage and/or surplus of prison guards, but two anomalies will be created. First, the results will be incompatible with equal-pay notions of fairness. If the unequal reluctance of males and females to enter this profession is considered by the evaluators, they will have to award more points to the female guards, since by stipulation they do the “same work”—this would be “unfair.”

Second, and more basically, if the evaluators take into account all phenomena that determine wages in the economy, of what possible use can they be? At best, they will no more than replicate the pattern of wages established on the marketplace. More likely, they will only imperfectly succeed in achieving this goal. After all, entrepreneurs succeed or fail in business to a great degree based on how closely they can tailor wage rates to productivity levels. The compensation of the “experts,” in contrast, will depend more on how well they satisfy their political constituencies. If there is, at best, only imperfect success in duplicating the market pattern of wages, this process will misallocate labor throughout the economy.

The implications for future research are clear. More attention should be paid to marital status as an explanation of female/male income differentials. Statistics should be published in a manner that more easily facilitates such research. Attempts should be made at an independent definition and measurement of discrimination. The residual method—especially in the face of non-employer discrimination—should be rejected. In comparing private- and public-sector discrimination, wage variances—not wage rates should be considered.

_____________________

Originally appeared in Toward Equity: Proceedings of a Colloquium on the Economic Status of Women in the Labor Market, Muriel Armstrong ed., Ottawa: The Economic Council, pp. 119–21, 179–82.

31 There are two major variants: equal pay for equal work and equal pay for work of equal value.

32 The debate between Brian Mulroney, John Turner, and Ed Broadbent on Women’s Issues, August 15, 1984. See The Globe and Mail and The Vancouver Sun, August 16, 1984. This is also the subject of bipartisan agreement in the United States, as the “comparable work” bill H.R. 5680 passed by a vote of 413 to 6 (on June 20, 1984).

33 See National Academy of Sciences, Women, Work, and Wages (Washington, D.C.: 1981). In this system, the “value” of jobs is determined by arbitrarily assigning points, or scores, to the standard elements of jobs such as skill, effort, responsibility, and working conditions.

34 Without wage differentials to lead workers around “as if by an invisible hand” (said Adam Smith), the only remaining market signal will be the varying unemployment rates in several industries. (Workers will shift from high-unemployment sectors to low-unemployment ones.) However, labor mobility will still be reduced, with only one effect, not two, working on its behalf. As well, there is a tendency for our present unemployment insurance program to enhance unemployment and to decrease labor mobility. See Herbert C. Grubel and Michael A. Walker (eds.), Unemployment Insurance (Vancouver: The Fraser Institute, 1978). It is thus difficult to see how the Canadian labor force can be successfully induced to organize itself in accordance with the ever-changing desires of Canadian consumers under a regime of relatively fixed wages.

35 It is clear that equal-pay enactments are but yet another version of wage controls. Yet, the experience of centuries of wage controls has shown that they misallocate labor, harm the best interests of most employees, and reduce economic efficiency. See Michael Walker, ed., The Illusion of Wage and Price Controls (Vancouver: The Fraser Institute, 1976).

36 Morley Gunderson, “Factors Influencing Male-Female Wage Differences in Ontario,” (Toronto: Research Branch, Ontario Ministry of Labour), p. 103. Similarly, in the Republic of South Africa, the white-racist unions advocate equal pay for equal work as a better means than job reservation laws of protecting their jobs against the competition of lower-paid black workers. See Leon Louw, “Free Enterprise and the South African Black,” Address to Barclay’s Executive Women’s Club, Johannesburg, South Africa, July 31, 1980, p. 4. This is a paradox. In South Africa, equal pay for equal work is advocated as a means of protecting a favored group (white unionists), while in Canada it is urged as a way of helping an unfavored group (females).

37 This data is for 1982 and applies to full-time employees. See Statistics Canada, Earnings of Men and Women, 1981 and 1982, Cat. 31-577, 1984.

38 Profit incentives, unfortunately, do not apply to the public sector. Bureaucrats who discriminate against women will thus run costlier operations, but they need not fear losing out to competitors and eventual bankruptcy. As a result, it would be plausible to expect a far greater rate of discrimination in government service than in the marketplace. Unfortunately, it is very difficult to make an empirical test of this proposition. This is because discrimination against women means paying them less than their marginal productivity, while paying men an amount equal to productivity or at least a wage rate closer to their productivity levels. The difficulty is that, in the public sector, there can be no independent measure of productivity. In the private marketplace, a wage rate indicates, at least in the ex ante sense, that the employer directly, and the consumer indirectly, values the contribution of the worker (productivity) more highly than the amount paid. In government service, in contrast, we are not even entitled to assume that marginal productivity always takes on a positive value. This is why in GNP accounts, government services are valued at cost; there are no independent measures of consumer valuation or productivity. A second problem is that it is by no means clear that only discrimination against women will take place in the public sector, nor even that discrimination against women will predominate over discrimination against men. Given the popularity of the idea of “reverse discrimination” among civil servants, it would be a heroic assumption to take for granted the direction of any discrimination that occurs.

Because the profit-and-loss system discourages discrimination in the marketplace but not in the public sector, The Fraser Institute recommended that “government efforts … ought to be directed primarily towards ensuring that discrimination does not occur in the public sector.” See Walter Block and Michael Walker, eds., “Preface,” Discrimination, Affirmative Action, and Equal Opportunity (Vancouver: The Fraser Institute, 1982), p. xvi. This recommendation has been challenged on the ground that female/male earnings ratios were actually higher in the public sector than in the private sector. See Margaret A. Denton and Alfred A. Hunter, Equality in the Workplace, Economic Sectors, and Gender Discrimination in Canada: A Critique and Test of Block and Walker … and Some New Evidence, Women’s Bureau, Discussion Paper Series A, No. 6: Equality in the Workplace (Ottawa: Labor Canada, 1982), pp. 12 and 34–40. But as we have seen, this finding is not incompatible with greater discriminatory behavior in the public sector—in this case against men and in favor of women. We know there will be more discrimination in the public sector than in the private sector. But we don’t know whether it will be against women, or men, or both. Thus, the only empirical implication of this hypothesis is a greater variance in wage rates (other things held constant), not a higher or lower female/male earnings ratio. But this test has yet to be made.

39Just as nature abhors a vacuum, economics abhors an unexploited profit opportunity. And the existence of “exploited” labor—that is, employees paid a wage below their productivity levels—is just such a profit opportunity. If men and women have equal productivity, say, $10 per hour, and men are paid $10 while women are paid $6.40, then the sex-neutral firm could hire all the women it wanted at, say, $7.00, and severely undercut its competition. But if there were competition between such sex-neutral employers, women’s wages would tend to be bid up to the $10 level. To illustrate just how powerfully business abhors a profit opportunity, consider the international mobility of multinational corporations. A large part of the motivation for the locational decisions of transnational enterprises is to “take advantage’ ‘of low-paid labor. They do so, of course, by moving in, opening a plant, and bidding wages up toward productivity levels.

40 Even Jane Fonda, whose radical feminist credentials need take a back seat to no one else’s, has run afoul of the equal-pay movement. Jane Fonda’s Workout is currently being sued for $3 million by three former female instructors on grounds of discrimination in pay compared with male employees who do essentially the same work. Ms. Fonda’s lawyer replies that the men were paid more because they were more productive in running clients through the exercise machines. See The Vancouver Sun, March 31, 1983, p. D8.

41 Thomas Sowell, Civil Rights: Rhetoric or Reality? (New York: William Morrow and Company, 1984), p. 97.

42 Barbara B. Reagan, “Two Supply Curves for Economists? Mobility and Career Attachment of Women,” American Economic Review 65, no. 2 (1975): 102.

43 Martin Meissner, “Sexual Division of Labor and Inequality: Labor and Leisure,” in Marylee Stephenson, ed., Women in Canada (Don Mills: General Publishing, 1977), pp. 166–74.

44 Sylvia Ostry, “Labor Force Participation and Child bearing Status,” Demography and Educational Planning (Toronto: Ontario Institute for Studies in Education, 1970), pp. 143–56.

45 Alan F. Bayer, “Marriage Plans and Educational Aspirations,” American Journal of Sociology 75 (1969): 239–44.

46 Single female academics who received their Ph.D.s in the 1930s outperformed their male counterparts by becoming full professors in the 1950s to a slightly greater extent, even though all women in academia fell far behind their male colleagues on a variety of indices. See Helen S. Astin, “Career Profiles of Women,” in Alice S. Rossi and Ann Calderwood, eds., Academic Women on the Move (New York: Russell Sage Foundation, 1973), p. 153 (cited in Thomas Sowell, “Weber and Bakke, and the Presuppositions of ‘Affirmative Action’,” in Block and Walker, eds., Discrimination).

47 According to Sowell, as of 1971, never-married females aged 30–39 earned more than never-married males of that age, even though all women as a group earned less than half of all men as a group. Could this be due to anti-male employer discrimination? See Thomas Sowell, Affirmative Action Reconsidered (Washington, D.C.: American Enterprise Institute, 1975), p. 28.

48 In the view of Denton and Hunter, in Equality in the Workplace, pp. 23–34, this effect of marriage on incomes cannot be substantiated empirically. However, their analysis is marred by a number of flaws.

First, they explicitly eschew our “narrow focus” on comparing never-married men and women; instead, they opt for a study of “all women in paid labor” (p. 24). But this leads to a poor test of our hypothesis. In our view, the best way to determine if there is any anti-female market discrimination is by considering only those who have never been touched by the institution of marriage: the never-married. A methodology that includes the ever-married (married, separated, divorced, widowed) cannot fully rule out the possibility of the asymmetric effects of marriage on male and female incomes. A better approach would have been to consider only the never-married category and to incorporate such independent variables of theirs as age, education, unionization, work interruptions, part-time and part-year work, and so on.

Second, they interpret discrimination as a residual; that is, they allot whatever earnings gap that cannot be explained by their other variables to discrimination. It is no more sensible to ascribe the pay differential between men and women to discrimination than it is to claim that the equally great or even greater income disparity between married and unmarried men is due to this source. Moreover, what of the possibility that this residual is due to other kinds of discrimination, apart from employer discrimination, such as discrimination against women on the part of consumers, fellow employees? Even if it made sense to interpret this residual as discrimination, it by no means logically follows that employer discrimination is necessarily responsible.

Third, even their own findings, improperly designed as they are, offer limited support to the marital-status hypothesis. Their Table 3 (p. 33) indicates that the “married” and “separated” categories are statistically significant at the 001 level. But why distinguish between “married,” “separated,” “divorced,” and “widowed?” A proper test of the marital-status hypothesis would make no such distinction; rather it would combine all these categories together into “ever-married” status.

Fourth, and even more problematical, the authors’ other independent variables serve as rough proxies for marital status (i.e., women with greater career interruptions are more likely to be married). Utilizing these, and then adding marital status, they find that these latter marriage categories add little or no explanatory power to the regression equations. But surely this is fallacious. For it was never contended that marital status makes a significant contribution to female/male income ratios over and above a whole host of variables that are correlated with marital status. This method of proceeding is almost guaranteed to invite problems of multicollinearity.

49 See Walter Block, “Economic Intervention,” in Block and Walker, eds., Discrimination, pp. 107–12. Alexander objects to this finding on the ground that the never-married “single women are … older than single men” and that age is related to productivity. See Judith A. Alexander, “Equal-Pay-for-Equal-Work Legislation in Canada,” Economic Council of Canada, Discussion Paper 252, (Ottawa: Economic Council of Canada, 1984), p. 27. But it is by no means clear that age is always positively related to productivity. On the contrary, it peaks in the mid-years; age is positively correlated with productivity before, negatively afterwards. In any case, the age difference in the sample was only 2.5 years, hardly enough to call into question an increase in the female/male income ratio of 374 for the total sample to 992 for the never-married. In this regard, perhaps their remark on p. 30 becomes more comprehensible: “It is impossible to refute Block’s arguments, although I do not find them convincing.” (Also, as reported in the original research, never-married females have 1.6 years more schooling than never-married males, work 3.3 more weeks in a year, and work on a part-time basis to a lesser degree, by 1.2 percent. However, since “living together” or “cohabitation” might be expected to have similar effects on male and female incomes as does marriage, and this phenomenon is not captured by official statistics on marital status, the ratio 992 may be an underestimation. Conceivably, the two biases might cancel each other out.)

50 So important is occupational segregation as an explanation of the female/male wage gap that it all but disappears when “productivity-adjusted comparisons are made within the same narrowly defined occupations within the same establishment—the wage gap that is most relevant for equal-pay legislation.” When this is done, the female/male adjusted earnings ratio “tends to be in the range of 90– 95.” See Morley Gunderson, The Female-Male Earnings Gap in Ontario: A Summary, Employment Information Series, no. 22 (Toronto: Ontario Ministry of Labor, February 1982), p. 17. Also, on the basis of occupational segregation, Walter Williams rejects the discrimination hypothesis as an explanation for the fact that black/white female income ratios (1.0225 in 1970) are much higher than black/white male income ratios (.6925 in 1970). See Walter Williams, “On Discrimination and Affirmative Action,” in Block and Walker, eds., Discrimination.

51 Sowell, in Civil Rights, p. 94, notes that “women have historically specialized in fields (such as editor, teacher, librarian) that they could leave and re-enter some years later, without large losses from obsolescence (as would occur as a tax attorney, aeronautical engineer, medical researcher).”

52 John M. McDowell, “Obsolescence of Knowledge and Career Publication Profiles: Some Evidence of Differences among Fields in Costs of Interrupted Careers,” American Economic Review 72, no. 4 (September 1982): 761.

53 Many feminist writers have claimed such phenomena do exist. See Meredith M. Kimball, “Women and Success: A Basic Conflict?” in Marylee Stephenson, ed., Women in Canada (Don Mills, Ont.: General Publishing, 1977), p. 85. For a more recent example of such commentary, see Judith Finlayson, “Any Way You Want Me,” Toronto Globe and Mail, October 12, 1984, p. 11.

54 C. Hoffman and J. Reed, “Imbalance Not Discrimination,” in Block and Walker, eds., Discrimination.

55 Michael Levin, “Comparable Worth: The Feminist Road to Socialism,” Commentary 78, no. 3 (September 1984): 15.

56 See, for example, Morley Gunderson, “Decomposition of the Male/Female Earnings Differential: Canada, 1970,” The Canadian Journal of Economics 12, no. 3 (August 1979): 479–85. See also Roberta Edgecombe Robb, “Earnings Differentials between Males and Females in Ontario, 1971,” The Canadian Journal of Economics 11, no. 2 (1978). For a reply to the latter, see Walter Block and Walter Williams, “Male-Female Earnings Differentials: A Critical Reappraisal,” The Journal of Labor Research 2, no. 2 (Fall 1981).

57 Sowell, in Civil Rights, pp. 46–47, says that:

One of the most important causes of differences in income and employment is the way people work—some diligently, carefully, persistently, cooperatively, and without requiring much supervision or warnings about absenteeism, tardiness, or drinking, and others requiring much concern over such matters. Not only are such things inherently difficult to quantify; any suggestion that such differences even exist is sure to bring forth a storm of condemnation. In short, the civil rights vision has been hermetically sealed off from any such evidence. Both historical and contemporary observations on inter-group differences in work habits, discipline, reliability, sobriety, cleanliness, or cooperative attitude—anywhere in the world—are automatically dismissed as evidence only of the bias or bigotry of the observers. “Stereotypes” is the magic word that makes thinking about such things unnecessary. Yet, despite this closed circle of reasoning that surrounds the civil rights vision, there is … evidence that cannot be disposed of in that way.

58 All attempts to discern objective values—whether for employment, goods, services, whatever—have failed, and have failed miserably. For critiques of the medieval theory of “just price,” and the Marxian attempt to establish “socially useful labor” as the objective measure of the value of goods and services, see Eugen von Böhm-Bawerk, “Value and Price,” Capital and Interest, book 3, vol. 2 (South Holland, Ill.: Libertarian Press, 1959).

59 We also assume, for the sake of argument, a necessarily segmented labor force; males cannot be hired to guard female prisoners, and females cannot be hired to guard male prisoners.

60 It by no means follows from the logic of the forced egalitarian philosophy that the lower wage of two occupations judged to be equal should be raised to match the higher one. It would be just as “equitable” to do the reverse. But most equal-pay legislation mandates raising the lower wage to the higher level. This creates unemployment. For in reality, if the marketplace wages of two groups are unequal, this tends to be based on unequal productivity levels. If the lower one is raised, elementary economic analysis indicates that unemployment will occur. See Walter Block, Focus on Economics and the Canadian Bishops (Vancouver, British Columbia: The Fraser Institute, 1983), pp. 45–55; and E.G. West and M. McKee, Minimum Wages: The New Issues in Theory, Evidence, Policy, and Politics, Economic Council of Canada (Ottawa: Supply and Services Canada, 1980).

• 1Philip Blumstein and Pepper Schwartz, American Couples: Money, Work, Sex (New York: William Morrow, 1983).
• 2See lewrockwell.com.
• 3See lewrockwell.com.
• 4“The Evil of Term Limits,” lewrockwell.com.
• 5See Hans-Hermann Hoppe, Democracy—The God That Failed: The Economics and Politics of Monarchy, Democracy, and the Natural Order (New Brunswick, N.J.: Transaction, 2001).
• 6See, e.g., Thomas Sowell, Race and Culture: A Worldview (New York: Basic Books, 1994).
• 7James Buchanan, Cost & Choice (Chicago: Markham Publishing, 1969), especially pp. 47–48; G.F. Thirlby, “Subjective Theory of Value and Accounting Cost,” Economica 13 (February 1946): 32–49; Ludwig von Mises, Human Action, 3rd ed. (Chicago: Regnery, 1963), pp. 242, 395.
• 8A female/male income ratio of 485 can be derived from Income Distributions by Size in Canada, 1979, Statistics Canada, Catalogue 13-207 (Ottawa: Statistics Canada, 1979), p. 99.
• 9Ratna Ray, Women in the Labour Force: Facts and Figures, Catalogue L 38-30/1977-2 (Ottawa: Department of Labour, Women’s Bureau, 1977), page i. Other economic studies of interest in this regard include Alan S. Blinder, “Wage Discrimination: Reduced Form and Structural Estimates,” Journal of Human Resources 8 (1973): 436–55; Morley Gunderson “Male-Female Wage Differentials and the Impact of Equal Pay Legislation,” Review of Economics and Statistics 57 (1975): 462–70; Morley Gunderson, “Work Patterns,” in Opportunity for Choice: A Goal for Women in Canada, ed. G. Cook (Ottawa: Statistics Canada, 1976).
• 10The average age of all employed Canadian men was 37 years in 1980. This placed the typical man in the highest male earnings age bracket (35–44 years old). The average age of all employed Canadian women was 34 years in 1980. This placed the typical woman in the fourth highest male earnings age bracket (25–34 years old). See The Labour Force, Statistics Canada, Catalogue 71-001, (Ottawa: Statistics Canada, 1980), page 75; and Income Distributions by Size in Canada, 1979, Statistics Canada, Catalogue 13-207 (Ottawa: Statistics Canada, 1979), pp. 104–09.
• 11Occupational segregation such as shown in the table below cannot properly be interpreted as the result of employer discrimination. After all, the employer cannot hire a nurse as a doctor, nor a person with secretarial training as an engineer or accountant.
• 12In 1978, the percentage of female Canadian union members was 28.7 percent; males comprised 71.3 percent of the membership. Source: Corporations & Labour Unions Returns Act, Part II, Labour Unions, Statistics Canada, Catalogue 71-202, p. 41. On the question of male/female productivity differentials, see Jacob Mincer and Solomon Polachek, “Family Investments in Human Capital: Earnings of Women,” Journal of Political Economy 82, no. 2, part 2 (March 1974): 76–108.
• 13In 1980, 94.1 percent of employed Canadian males worked full time, 5.9 percent worked part time; 76.2 percent of employed Canadian females worked full time, 23.8 percent worked part time. In 1979, 72.9 percent of employed Canadian males worked 50–52 weeks, 27.1 percent worked 1–49 weeks; 60.1 percent of Canadian females worked 50–52 weeks, 39.9 percent worked 1–49 weeks. Source: The Labour Force, Statistics Canada, December 1980, p. 105, Catalogue 71-001; Income Distributions by Size in Canada, 1979, Statistics Canada, Catalogue 13-207, p. 112.
• 14These figures are derived from computations based on data cited for elementary school teachers’ earnings and percentages of teachers by sex: Salaries and Qualifications of Teachers in Public, Elementary & Secondary Schools, 1979–80, Statistics Canada, Catalogue 81-202, p. 25; university teachers’ earnings and percentages of university teachers by sex: Teachers in Universities, 1978–1979, Statistics Canada, Catalogue 81–241, p. 57. Note that while we have corrected the female/male income ratio for several phenomena not related to discrimination, a still more accurate assessment would have to normalize for all of these variables, together, and include other variables such as continuity of employment, earned degrees, labor force participation, location, industrial concentration, public or private employment, productivity, seniority, as well as such imponderables as motivation, “stick-to-it-iveness,” resourcefulness, ambition, expectations, etc.
• 15Thomas Sowell, Affirmative Action: Reconsidered (Washington, D.C.: American Enterprise Institute, 1975), pp. 23–34.
• 16Jesse Bernard, Academic Women (University Park: Pennsylvania State Univ. Press, 1964), pp. 220–26; Jesse Bernard, The Future of Motherhood (New York: Penguin Books, 1974), pp. 165–70; Bryan and Boring, American Psychologist 2 (January 1947): 18; Lee Rainwater, And the Poor Get Children (Chicago: Quadrangle Books, 1960), pp. 67–69; Wayne R. Bartz and Richard A. Rasor, Surviving With Kids (San Luis Obispo, Calif.: Impact, 1978), p. 147; Martin Meissner, “Sexual Division of Labour and Inequality: Labour and Leisure,” in Women in Canada, pp. 166–74; Nancy Chodorow, “Being and Doing: A Cross Cultural Examination of the Socialization of Males and Females,” in Women in Sexist Society, ed. Vivian Gornick and Barbara K. Moran (New York: Basic Books, 1971), pp. 183–84; Roslyn S. Willett, “Working in ‘A Man’s World’: The Woman Executive,” in Women in Sexist Society, p. 368; Jean Tepperman, “Two Jobs: Women Who Work in Factories,” in Sisterhood is Powerful, ed. Robin Morgan (New York: Random House, 1970), pp. 115, 121.
• 17Gail C.A. Cook, “Opportunity for Choice: A Criterion,” in Opportunity for Choice: A Goal for Women in Canada, ed. Gail C.A. Cook (Ottawa: Statistics Canada and C.D. Howe Research Institute), Catalogue IC 23-15/1976, p. 4; Gail C.A. Cook and Mary Eberts, “Policies Affecting Work,” in Opportunity for Choice, p. 145; Richard A. Lester, Antibias Regulations of Universities (New York: McGraw-Hill, 1974), p. 39; Willett, “Working in ‘A Man’s World’,” p. 368; Pat Mainardi, “The Politics of Housework,” in Sisterhood Is Powerful, pp. 447–54; Jesse Bernard, The Future of Motherhood, pp. 157–65; Kathryn E. Walker, “Time Used by Husbands for Household Work,” Family Economics Review (June 1970): 8–10; M. Meisner, E.W. Humphries, S.M. Meis, and W.J. Scheu, “No Exit for Wives: Sexual Division of Labour and the Cumulation of Household Demands,” Canadian Review of Sociology and Anthropology 12 (1975): 424–39.
• 18Barbara B. Reagan, “Two Supply Curves for Economists? Implications of Mobility and Career Attachment of Women,” American Economic Review 65, no. 2 (1975): 102; Jacquelyn S. Crawford, Women in Middle Management (Ridgewood, N.J.: Forkner, 1977), p. 63.
• 19E.W. Burgess and Paul Wallin, Engagement and Marriage (New York: Lippincott, 1953), pp. 614, 618; and Reagan, “Two Supply Curves,” p. 104. See also Beth Neimi, “The Female-Male Differential in Unemployment Rates,” Industrial and Labour Relations Review 27, no. 3 (April 1974): 331–50.
• 20Alan E. Bayer, “Marriage Plans and Educational Aspirations,” American Journal of Sociology 75 (1969): 239–44; Reagan, “Two Supply Curves,” p. 103; Bernard, The Future of Motherhood, pp. 91, 151, 181; Jean Tepperman, “Two Jobs,” p. 123; Betty Friedan, The Feminine Mystique (New York: Dell, 1974), p. 31; Meg Luxton, More Than a Labour of Love: Three Generations of Women’s Work in the Home (Toronto: Women’s Educational Press, 1980), p. 16; Margaret Luxton, “Urban Communes and Co-ops in Toronto,” (M. Phil. dissertation, University of Toronto, 1973), cited in Luxton, More Than a Labour of Love; Ann Oakley, Women’s Work: The Housewife Past and Present (New York: Vintage Books, 1976); Eli Zaretsky, Capitalism, The Family, and Personal Life (New York: Harper & Row, 1976), p. 17; Simone de Beauvoir, The Second Sex (New York: Vintage Books, 1974), p. 482.
• 21Let us imagine an experiment. We offer a large number of employed married couples the following option: jobs in city A, where the husband will earn $200,000 per year and the wife $150,000, or in city B, where the wife will earn $200,000 per year, and the husband $150,000. (The type of employment and the amenities of the cities are assumed to be identical in each case.) How many of the husbands would prefer city B? How many wives would prefer city A? Although there is only casual evidence on this, since such an experiment has not yet been done, one may speculate that there will be more wives who will prefer city A than there will be husbands who will prefer city B. The motivations behind these choices may vary. Some husbands may feel “less of a man” if their wives earn more than they do; others may feel it is just “not fitting” that their incomes should be lower; some wives may feel less damaged psychologically from earning less than their spouses; others may subscribe to the societal pressures which teach, at a young age, that “nice girls don’t outcompete boys.” But whatever the reason, there is abundant anecdotal evidence that many women have great psychological and other personal difficulties in competing with men, and are thus, when married, more likely than men to purposefully keep their earnings below those of their spouses—with important implications for the low female/male earnings ratios for married people.
• 22Unfortunately, only the “married” and “total” columns are widely published in official Canadian Statistics. To say the least, this gives rather a biased account of the true state of the female/male earnings ratio, and its basic cause (marital status and its widely diverging effects on male and female earnings).
• 23Not only does marriage have asymmetrical effects on earnings by sex, but it is reasonable to believe that so does “living together” or “cohabitation”—and for similar reasons. Moreover, this category has become more statistically significant in recent years, though actual data are lacking. Table 3, in distinguishing between the ever married and the never married, cannot separate those who have ever lived together—whether married or unmarried—from those who have never lived together. If cohabitation as well as ever-married status could be controlled for, one would thus expect the female/male ratio to be higher than 992. There is also, however, a reason for believing that 992 may be somewhat of an overestimate of the “true” female/male earnings ratio: never-married females are older than never-married males (46.2 years vs. 43.7 years old), have more schooling (10.9 vs. 9.3 years), work more weeks (45.6 vs. 42.3 weeks), and work on a part-time basis to a lesser degree (10.6 percent vs. 11.8 percent). (All figures based on calculations made from data presented by Kuch and Haessel; see Table 3 for full citation.)
• 24Some may argue that private discriminatory behavior is immoral and ought to be prohibited by force of law. Others may hold that, while discrimination is a negative characteristic, each individual is nevertheless entitled to make whatever decision suits his conscience—whether in commercial dealings, employment practices, housing decisions, or personal relations—provided only that he not commit fraud upon, or initiate aggression against, minority group members. Whatever the solution to this philosophical question, both sides may perhaps take comfort from one of the findings in Block and Walker, Discrimination, Affirmative Action, and Equal Opportunity: the tendency of the marketplace to financially penalize those who indulge in discriminatory practices, and thus to reduce, over time, the scope of this activity.
• 25There would appear to be numerous criteria upon which discrimination has, or is alleged to have, taken place, and upon which quotas, affirmative action, or preferential treatment are now demanded. Some of the grounds include (1) obesity: see “Obese Are Victims of Bias: Professor,” Toronto Globe & Mail (August 5, 1980), p. 13; (2) blindness: see “Group for Blind Suggests Job Quota,” Toronto Globe & Mail (August 2l, 1980), p. 1, and “Blind Woman and Guide Dog Win Rights Fight,” Vancouver Sun (August 18, 1980), p. 8; (3) residence: see “Stop Provinces Reserving Jobs for Residents, Rights Chief Says,” Toronto Globe & Mail (June 4, 1980), p. 10; (4) “reverse” discrimination: see “Barring White in Native Class Is Ruled Illegal,” Toronto Globe & Mail (February 13, 1980), which tells of an Alberta Human Rights Commission finding against the University of Calgary for discriminating against a non-Indian woman by denying her admission to a special course specifically set up for native peoples; (5) ugliness: see “More to an Interview than Meets the Eye,” Toronto Globe & Mail (July 19, 1980), p. F3, which shows that persons of “unattractive appearance” (and even sometimes persons of beauty) are discriminated against in hiring practices; (6) political beliefs: see “They’re Biting the Hand that Won’t Feed Them,” Toronto Globe & Mail (August 9, 1980), p. 8, which tells of a Prince Edward Island Provincial Human Rights Commission finding that a public employee had been wrongfully dismissed for his political beliefs; (7) hirsuteness: see “Supreme Court Refuses a Motion to Force Grocery Clerk to Shave,” Toronto Globe & Mail (February 10, 1980), which tells of a Winnipeg employer who could not legally compel his grocery clerk to either shave his beard, work nights “out of the sight of customers,” or fire him.
• 26The situation with regard to discriminatory practices on the part of the government is a unique matter. It cannot be argued, as it could in the private case, that, no matter how morally reprehensible discrimination is, at least in the market sector it is done by an individual in his own name and with his own money.
• 27Although many people interpret prejudice or discriminatory behavior as a willingness to engage in physical aggression against a despised person or group, this interpretation is about as far away as it is possible to get from a clear understanding. On the contrary, a sharp distinction must be made between refusing to interact at all with a person, and threatening physical abuse against him. The former is all that is done by a hermit, although on a larger scale; and if it is no crime to refuse to deal with all of humanity, then it might be argued that it can scarcely be a rights violation to avoid dealing with only some people.
• 28Ludwig von Mises, Human Action, p. 109.
• 29Toronto Globe & Mail (April 14, 1980), p. 6. See also Toronto Globe & Mail (January 25, 1980).
• 30See “Drea Tells Firm to Stop Questions” and the editorial “Wearing Nothing but a Seat Belt,” both in Toronto Globe & Mail (July 31, 1980), pp. 5, 6.
• 31Most women, on the other hand, even executives, do not consider housework demeaning.
• 32In the first of these columns, “A Silver Lining in Drug Prohibition,” I noted that real criminals who might not otherwise be caught are sometimes jailed for engaging in the drug trade, which they have every right to do. But fewer murderers and rapists at large is surely a good thing. In the second of these columns, “A Silver Lining in Unjust Executions,” I observed that those executed for murders for which they are innocent (a travesty of justice if ever there was one) were sometimes guilty of committing other capital crimes for which they were not charged. Thus, a critique of the death penalty was not as powerful as otherwise thought. In the third, I tried to clarify my position on this issue vis-à-vis several objections that had been made of it.
• 33In the first of these columns, “A Silver Lining in Drug Prohibition,” I noted that real criminals who might not otherwise be caught are sometimes jailed for engaging in the drug trade, which they have every right to do. But fewer murderers and rapists at large is surely a good thing. In the second of these columns, “A Silver Lining in Unjust Executions,” I observed that those executed for murders for which they are innocent (a travesty of justice if ever there was one) were sometimes guilty of committing other capital crimes for which they were not charged. Thus, a critique of the death penalty was not as powerful as otherwise thought. In the third, I tried to clarify my position on this issue vis-à-vis several objections that had been made of it.
• 34In order for there to be a silver lining, there has to be an evil, and, also, some good must come out of it. I am already on record in taking the position that term limits are highly problematic. I did so on grounds blazed by Professor Hans Hoppe. His argument is based on time preference: other things equal, the longer a politician is in office, the more of a long-run viewpoint he can afford to take. If he can leave the office to his children (e.g., a monarchy), he will act even more responsibly; he doesn’t want to kill the golden goose, otherwise there will be nothing left for his progeny to exploit.
• 35In order for there to be a silver lining, there has to be an evil, and, also, some good must come out of it. I am already on record in taking the position that term limits are highly problematic. I did so on grounds blazed by Professor Hans Hoppe. His argument is based on time preference: other things equal, the longer a politician is in office, the more of a long-run viewpoint he can afford to take. If he can leave the office to his children (e.g., a monarchy), he will act even more responsibly; he doesn’t want to kill the golden goose, otherwise there will be nothing left for his progeny to exploit.
• 36There is no one who has done more to combat this pernicious fallacy than Tom Sowell. But he needs all the help he can get on this mission. It is incumbent upon all men of good will to help him in this regard. And one way that we can do so is to recognize one of the side order benefits of that otherwise insidious policy of term limits: that it disproportionately penalizes female politicians.
• 37Another difficulty is that “equal opportunity” is a subjective, not an objective, phenomenon. Women do not come equipped with a little tag which indicates their productivity, once and for all, in a manner from which no dissent is possible. (Nor, of course, do men.) Productivity, rather, is a continually changing phenomenon which varies with, for example, education, intelligence, age, experience, presence or absence of complementary factors of production, which can only be partially quantified, as well as with such factors as motivation and determination, which are completely incapable of exact measurement.
• 38This is not to say it is completely implausible to support the position that EPFEW laws are required. There is ample evidence to suggest that male and female compensation and promotion rates differ. For example, the overall female/male wage ratio for all employees in Canada was 485 in 1979. That is, females, on average, earned only 48.5 percent as much as males. The question is whether or not this reflects an inherent problem in the labor market. The proponents of EPFEW apparently think so. But it is an error to conclude, from such information, that the state of affairs is a result of conscious and hostile human design (i.e., employer discrimination against women), that this is disgraceful and unfair, and that therefore a determined effort on the part of government is required to “right these wrongs.” Consider, in this regard, the statement of no less an authority than Dr. Ratna Ray, Director, Women’s Bureau, Labour Canada:
• 39Economic self-sufficiency for women? “We’ve only just begun!” Despite sporadic progress, there’s a long way to go. Canadian employers, economic planners, and decision-makers should take a long hard look at these figures. Because they tell a shocking story, a story of shortsightedness and languorous efforts towards the utilization of human resources in the workforce. Our publication comes hard on the heels of the National Council of Welfare’s report Women and Poverty which concludes that “After fifty years or so of unpaid, faithful service a woman’s only reward is likely to be poverty.”
• 40Men and women differ in a wide range of economic, educational, and sociological characteristics, each of which exerts an independent effect, raising expected male incomes and reducing expected female incomes. For example, working men tend to be older than women, more highly concentrated in the higher-paying professions, more heavily unionized in the highly skilled and legislatively protected blue-collar industries, and tend to work, to a greater degree, on a full-time, full-year basis.
• 41Men and women differ in a wide range of economic, educational, and sociological characteristics, each of which exerts an independent effect, raising expected male incomes and reducing expected female incomes. For example, working men tend to be older than women, more highly concentrated in the higher-paying professions, more heavily unionized in the highly skilled and legislatively protected blue-collar industries, and tend to work, to a greater degree, on a full-time, full-year basis.
• 42Men and women differ in a wide range of economic, educational, and sociological characteristics, each of which exerts an independent effect, raising expected male incomes and reducing expected female incomes. For example, working men tend to be older than women, more highly concentrated in the higher-paying professions, more heavily unionized in the highly skilled and legislatively protected blue-collar industries, and tend to work, to a greater degree, on a full-time, full-year basis.
• 43Men and women differ in a wide range of economic, educational, and sociological characteristics, each of which exerts an independent effect, raising expected male incomes and reducing expected female incomes. For example, working men tend to be older than women, more highly concentrated in the higher-paying professions, more heavily unionized in the highly skilled and legislatively protected blue-collar industries, and tend to work, to a greater degree, on a full-time, full-year basis.
• 44Not unexpectedly, when female/male income comparisons have been corrected for these factors, the ratio tends to rise. If working women are assumed to retain their own income levels, but to take on the same age pattern as working men, the female/male income ratio rises from 485 to 521; if we assume that females are divided among elementary and university teaching in the same manner as males, the ratio increases from 743 to 814; if the female/male income ratio is corrected in a similar manner for full-year or part-year status, it increases from 528 to 575. See Table l.
• 45As important as these variables are, the strongest determinant of the so-called male/female earnings “gap” is none of these things. Rather, it is marital status, and the asymmetric effects of marriage on male and female earnings. That is, marriage increases male earnings, and reduces female earnings.
• 46This occurs for many reasons. Wives, to a greater degree than husbands, take on a higher and disproportionate share of child care and the housework and homemaker tasks. They are more likely to quit their jobs if their partner receives a better job elsewhere, to interrupt their careers for domestic reasons, to place their homes and families ahead of their jobs or professions, and even to purposefully attempt to keep their earnings below that of their spouses. It is impossible to quantify the effects of such phenomena in driving a wedge between married male and female incomes, but there is little doubt that they are important.
• 47This occurs for many reasons. Wives, to a greater degree than husbands, take on a higher and disproportionate share of child care and the housework and homemaker tasks. They are more likely to quit their jobs if their partner receives a better job elsewhere, to interrupt their careers for domestic reasons, to place their homes and families ahead of their jobs or professions, and even to purposefully attempt to keep their earnings below that of their spouses. It is impossible to quantify the effects of such phenomena in driving a wedge between married male and female incomes, but there is little doubt that they are important.
• 48This occurs for many reasons. Wives, to a greater degree than husbands, take on a higher and disproportionate share of child care and the housework and homemaker tasks. They are more likely to quit their jobs if their partner receives a better job elsewhere, to interrupt their careers for domestic reasons, to place their homes and families ahead of their jobs or professions, and even to purposefully attempt to keep their earnings below that of their spouses. It is impossible to quantify the effects of such phenomena in driving a wedge between married male and female incomes, but there is little doubt that they are important.
• 49This occurs for many reasons. Wives, to a greater degree than husbands, take on a higher and disproportionate share of child care and the housework and homemaker tasks. They are more likely to quit their jobs if their partner receives a better job elsewhere, to interrupt their careers for domestic reasons, to place their homes and families ahead of their jobs or professions, and even to purposefully attempt to keep their earnings below that of their spouses. It is impossible to quantify the effects of such phenomena in driving a wedge between married male and female incomes, but there is little doubt that they are important.
• 50This occurs for many reasons. Wives, to a greater degree than husbands, take on a higher and disproportionate share of child care and the housework and homemaker tasks. They are more likely to quit their jobs if their partner receives a better job elsewhere, to interrupt their careers for domestic reasons, to place their homes and families ahead of their jobs or professions, and even to purposefully attempt to keep their earnings below that of their spouses. It is impossible to quantify the effects of such phenomena in driving a wedge between married male and female incomes, but there is little doubt that they are important.
• 51This occurs for many reasons. Wives, to a greater degree than husbands, take on a higher and disproportionate share of child care and the housework and homemaker tasks. They are more likely to quit their jobs if their partner receives a better job elsewhere, to interrupt their careers for domestic reasons, to place their homes and families ahead of their jobs or professions, and even to purposefully attempt to keep their earnings below that of their spouses. It is impossible to quantify the effects of such phenomena in driving a wedge between married male and female incomes, but there is little doubt that they are important.
• 52() Married and Total figures are published in each year’s Income Distribution by Size (Statistics Canada, Catalogue 13-207). Single and Other averages are from unpublished tables. Survey of Consumer Finances, Statistics Canada.
• 53In order to remedy this situation, we turn to Table 3, which includes only wages and salaries and specifically excludes all other income. As a test of the hypothesis that marital status has widely asymmetrical effects on earnings by sex, Table 3 is an improvement over Table 2 in yet another way: it collapses the categories of “married” and “other” into “ever-married,” which, as the name implies, includes all people who were ever-married, regardless of the marital status they now occupy. That is, it compares people presently married, divorced, widowed, or separated, with those who were never married in their entire lives. Table 3 thus furnishes a comparison of people who have been touched in some way by the institution of marriage with those who have not.
• 54Given the great difficulties, social costs, and unintended negative consequences of proportional representation requirements, quotas, and other similar prescriptions based on retrospective results, and given the market’s ability to reduce discriminatory behavior—in the absence of legislation which retards this process—a case is made, throughout this book, against the employment of affirmative action programs in the private sector. We also noted above that the incentive system operating in the private sector may not work in the public sector. Accordingly, we must now assess the case for equal opportunity programs (based on quotas and proportional representation) in the field of public employment.
• 55Discrimination in the public sector, moreover, is considered unjust and immoral by many. The funds which pay for government employment come from all citizens—including minority members. For anyone to be excluded from a public job because of race, sex, national origin, sexual preference, skin color, or any other such criteria, after being forced to pay for this very same unemployment through coercive taxes, is nothing more than a cleverly disguised, but particularly insidious, form of exploitation.
• 56Discrimination in the public sector, moreover, is considered unjust and immoral by many. The funds which pay for government employment come from all citizens—including minority members. For anyone to be excluded from a public job because of race, sex, national origin, sexual preference, skin color, or any other such criteria, after being forced to pay for this very same unemployment through coercive taxes, is nothing more than a cleverly disguised, but particularly insidious, form of exploitation.
• 57The difficulty is that quotas are unjust. The beneficiaries (in those rare cases where someone actually benefits) are the wrong people: the 18- to 21-year-olds, applying for their first jobs, who never bore the brunt of past employment discrimination, by definition. The real victims are those who would have liked to have been police officers, firefighters, letter carriers, and civil servants in the past but were not even considered, even though fully qualified, because of their race or sex. But these people, for the most part, are already either settled in their present jobs or retired from the labor force. If anything, monetary settlements might seem a preferable alternative.
• 58Insurance is an industry dedicated to pooling and spreading risk. While health, injury, or sickness of any one person is beyond prediction, actuarial tables have been established for the probability of such occurrences in the aggregate. Because of this, insurance companies can charge premiums to large numbers of people and underwrite the costs of the unfortunate few; all customers pay a relatively small amount, in effect, for the security that should they be struck by calamity, they or their loved ones would be protected from great expenses, and thus would remain financially secure.
• 59A case in point is the recent Canadian Human Rights Commission regulation condemning discrimination between men and women with regard to pension plans. It is a plain actuarial fact—established through years of insurance experience—that women tend to live longer than men. With sexual discrimination prohibited, equal pension premiums would render men more profitable customers to insurance companies, as on average they will collect benefits for fewer years. The Canadian Human Rights Commission prohibition will therefore tend to result in (1) the encouragement of male over female labor (men will now be cheaper to employ); (2) the segregation of the labor force by sex (so that no one employer would have to make different contributions on this basis); (3) the withdrawal of employers, especially small ones, from pension plans altogether; and/or (4) the migration of companies to areas which do not prohibit discrimination in pension premium payments on the basis of sex. Needless to say, any of these eventualities would effectively discourage workers from pooling risks concerning retirement income.
• 60Another case in point is the order of the Ontario Consumer and Commercial Relations Ministry to the Fireman’s Fund Insurance Company. The ministry enjoined the insurance company from questioning its clients about their convictions on alcohol or sex-related offences, on whether they are usually restless, sad, or sweaty at night, and on other personal subjects.