Labor Market Discrimination 1 Hwei-Lin Chuang, Ph.D. 2010/05/27

網路求才 限男限女無法管 保障民眾平等就業機會，法律禁止以性別、 年齡、性傾向等作為求職限制。但法律只規 範雇主。 【聯合報╱記者黃驛淵／台北報導】 2010.05.16 保障民眾平等就業機會，法律禁止以性別、 年齡、性傾向等作為求職限制。但法律只規 範雇主。 許多人力網站、家教網、知名BBS「PTT實業 坊」等網路平台都充斥「限女、限男」，「 女佳、男佳」等涉嫌歧視字眼卻無法可管。 勞工局表示，法律僅規範「資方及雇主」， 對刊載的媒體平台包括網路的確無法可罰， 僅能行文要求註明法律相關責任。

網路求才 限男限女無法管 去年清查網路徵才廣告時，發現各人力網及 PTT上有類似廣告，已要求在版上敘明就業 服務法及性別工作平等法規定。 勞工局表示，雇主刊登求職廣告若出現限女 、限男已違反就業歧視、性別平等。 可依就業服務法處30萬元以上150萬元以下罰鍰。 民眾只要遇到任何求職歧視，都可提出申訴。 資料來源﹕ http://udn.com/NEWS/DOMESTIC/DOM2/5602307.shtml

相關法例 就業服務法第五條: 為保障國民就業機會平等，雇主對求職人或所僱用員工 ，不得以種族、階級、語言、思想、宗教、黨派、籍貫 、性別、容貌、五官、殘障或以往工會會員身分為由， 予以歧視。 資料來源﹕ http://www.labor.gov.tw/web/j9.htm

相關法例 性別工作平等法 (節錄第二章性別歧視之禁止) 第 7 條　雇主對求職者或受僱者之招募、甄試、進用、分發、 配置、考績或陞遷等 ，不得因性別或性傾向而有差別待遇。但 工作性質僅適合特定性別者，不在此限。 第 8 條　雇主為受僱者舉辦或提供教育、訓練或其他類似活動 ，不得因性別或性傾 向而有差別待遇 第 9 條　雇主為受僱者舉辦或提供各項福利措施，不得因性別 或性傾向而有差別待 遇。 第 10 條　雇主對受僱者薪資之給付，不得因性別或性傾向而有 差別待遇；其工作或 價值相同者，應給付同等薪資。但基於年 資、獎懲、績效或其他非因性別 或性傾向因素之正當理由者， 不在此限。 雇主不得以降低其他受僱者薪資之方式，規避前項 之規定。

相關法例 第 11 條　雇主對受僱者之退休、資遣、離職及解僱，不 得因性別或性傾向而有差別 待遇。 工作規則、勞動契約 或團體協約，不得規定或事先約定受僱者有結婚、懷 孕 、分娩或育兒之情事時，應行離職或留職停薪；亦不得 以其為解僱之理 由。 違反前二項規定者，其規定或約定 無效；勞動契約之終止不生效力。 資料來源﹕ http://law.moj.gov.tw/LawClass/LawAll.aspx?PCode=N0030 014

Labor Market Discrimination Differences in earnings and employment opportunities may arise even among equally skilled workers employed in the same job simply because of the workers’ race, gender, national origin, sexual orientation, and other seemingly irrelevant characteristics. These differences are often attributed to labor market discrimination.

OUTLINE The Discrimination Coefficient Employer Discrimination Employee Discrimination Consumer Discrimination Statistical Discrimination Measuring Discrimination

表1. 受雇者每月工作薪資-按年齡別分 主計處人力運用調查報告 表1. 受雇者每月工作薪資-按年齡別分 主計處人力運用調查報告 項目別 兩性平均 女性工作收入佔男性收入比率 年齡別 \ 年度 94 95 96 97 98 １５～２４歲 22,950 23,175 23,372 22,346 20,632 97.65% 94.25% 96.04% 95.71% 100.20% １５～１９歲 17,416 16,820 17,245 16,611 14,875 97.02% 87.53% 91.71% 91.24% 93.25% ２０～２４歲 23,850 24,201 24,415 23,457 21,685 95.76% 93.11% 95.02% 94.79% 101.39% ２５～４４歲 34,047 34,158 34,272 34,457 33,441 80.27% 80.97% 80.49% 80.32% 81.94% ２５～２９歲 29,554 29,866 30,047 29,806 28,669 88.75% 90.73% 90.62% 90.23% 89.61% ３０～３４歲 34,139 34,022 33,970 33,995 32,787 81.66% 81.32% 82.81% 83.33% 84.20% ３５～３９歲 36,114 36,192 36,105 36,771 35,739 78.59% 79.96% 76.69% 78.70% 81.40% ４０～４４歲 37,443 37,583 37,932 38,251 37,847 75.18% 74.43% 74.67% 72.52% 75.63% ４５～６４歲 37,522 37,564 37,805 38,885 38,878 70.08% 69.72% 70.16% 70.82% 72.12% ４５～４９歲 37,860 37,550 38,236 39,346 38,501 71.68% 69.43% 73.39% 71.42% 72.50% ５０～５４歲 36,907 37,418 36,641 39,302 39,142 70.76% 71.89% 67.00% 71.32% 74.76% ５５～５９歲 38,964 38,766 39,645 37,499 39,397 65.21% 67.63% 66.42% 68.64% 69.45% ６０～６４歲 33,073 34,120 34,943 35,820 38,624 59.81% 61.67% 80.01% 61.46% 59.18% ６５歲及以上 24,529 23,054 25,089 31,135 26,615 44.10% 48.39% 45.65% 47.51% 58.66%

表2. 受雇者每月工作薪資-按教育程度別分主計處人力運用調查報告 項目別 兩性平均 女性工作收入佔男性收入比率 教育程度 \ 年度 94 95 96 97 98 國中及以下 25,957 26,021 26,510 26,897 25,698 68.95% 69.71% 69.88% 69.28% 70.42% 國小及以下 25,519 24,993 25,975 23,644 66.35% 68.61% 68.74% 67.36% 67.71% 國中 26,348 26,680 26,845 27,473 26,705 71.37% 71.47% 71.14% 71.21% 73.00% 高中（職） 29,850 29,840 29,957 29,913 29,065 76.37% 76.85% 75.14% 75.65% 78.07% 高中 30,416 29,981 30,785 30,539 29,789 75.17% 75.63% 73.66% 73.96% 77.62% 高職 29,669 29,787 29,689 29,707 28,845 77.25% 75.79% 76.29% 78.28% 大專及以上 41,274 41,115 41,072 41,223 39,933 75.74% 76.07% 74.99% 75.99% 專科 36,516 36,314 36,425 37,263 36,114 76.02% 78.95% 77.09% 75.26% 76.61% 大學及以上 45,869 45,286 44,714 44,027 42,388 76.30% 74.29% 75.75% 75.10%

1. The Discrimination Coefficient Money, commonly used as a measuring rod, will also serve as a measure of discrimination. →If an individual has a " taste for discrimination," he must act as if he were willing to pay something, either directly or in the form of a reduced income, to be associated with some persons instead of others. →By using the concept of a discrimination coefficient (DC), it is possible to give a definition of a " taste for discrimination." It is parallel for different factors of productions, employers, and consumers.

A DC represent a nonpecuniary element in certain kinds of transactions, and it is positive or negative, depending on whether the nonpecuniary element is considered "good" or "bad". →Discrimination is commonly associated with disutility caused by contact with some individuals.

Discrimation : di , dj ,dk > 0 Nepotism : di , dj ,dk < 0 Money cost(return) net cost(return) Employer π π (1+di) Employee πj πj (1-dj) Consumer P P× (1+dk) Discrimation : di , dj ,dk > 0 Nepotism : di , dj ,dk < 0 (親戚主義，裙帶關係) →This quantitative representation of a taste for discrimination provides the means for empirically estimating the quantitative importance of discrimination.

2. Employer Discrimination There are two types of workers in the labor market: white workers and black workers. We assume that black and white workers are perfect substitutes in production, so that the production function can be written as: (8-1) Where q is the firm’s output, EW gives the number of white workers hired, and EB gives the number of black workers hired. A firm that does not discriminate will hire black workers up to the point where the black wage equals the value of their marginal product, or: (8-2)

Employment in a Discriminatory Firm The employer acts as if the black wage is not WB, but is instead equal to WB× (1+d), where d is the discrimination coefficient. The decision rule for an employer that discriminates against blacks is: Hire only blacks if WB× (1+d)<WW Hire only whites if WB× (1+d)>WW (8-3) → As long as black and white workers are prefect substitutes, firms have a segregated work force.

There are, therefore, two types of firms: “white firms,” and “black firms.” Employers who have small discrimination coefficients will hire only blacks; employers with large discrimination coefficients will hire only whites. Dollars Employment VMPE WW EW* (a) White Firm (b) Black Firm Figure 1. The Employment Decision of a Prejudiced Firm WB WB(1+d1) WB(1+d0) EB* EB1 EB0

Discrimination and Profits The relationship between the firm’s profits and the discrimination coefficient is illustrated in Figure 2. The most profitable firm has a zero discrimination coefficient and has profits of πmax dollars. This color- blind firm hires an all-black work force and EB* workers. Firms with slightly positive discrimination coefficients still have an all-black work force, but employ fewer black workers and earn lower profits.

At some threshold level of prejudice, given by the discrimination coefficient dW, the utility loss of hiring blacks is too large and the firm hires only whites. As a result, profits take a dramatic plunge (toπW dollars) because the firm is paying a much higher wage than it needs to. Because all white firms hire the same number of white workers (or EW*) regardless of their discrimination coefficient, all white firms earn the same profits.

Firms that discriminate lose on two counts: They are hiring the “wrong color” of workers and/or they are hiring the wrong number of workers. Figure 2. Profits and the Discrimination Coefficient Dollars Black Firms White dW πmax πW Discrimination Coefficient

3. Employee Discrimination The source of discrimination in the labor market need not be the employer, but might instead be fellow workers. Suppose that whites dislike working alongside blacks and that blacks are indifferent about the race of their coworkers. As we see, white workers who receive a wage of WW dollars will act as if their wage rate is only Ww × (1-d), where d is the white worker’s discrimination coefficient.

A color-blind profit-maximizing employer would never have an integrated work place. The employer would not hire both black and white workers because white workers have to be paid a compensating wage differential, yet they have the same value of marginal product as black. Hence, the employer will hire only whites if the white wage is below the black wage, and will hire only blacks if the black wage is below the white wage.

Unlike employer discrimination, however, employee discrimination does not generate a wage differential between equally skilled black and white workers. Color-blind employers hire whichever labor is cheaper. Note that employee discrimination does not affect the profitability of firms. Because all firms pay the same price for an hour of labor, and because black and white workers are perfect substitutes, there is no advantage of being either a black or a white firm.

4. Consumer Discrimination If consumers have a taste for discrimination, their purchasing decisions are not based on the actual price of the good, p, but on the utility-adjusted price, or p × (1+d),where d is the discrimination coefficient. As long as a firm can allocate a particular worker to one of many different positions within the firm, consumer discrimination may not matter much. If the firm cannot easily hide black workers from public view, however, consumer discrimination can have an adverse impact on black wages. A firm employing a black worker in a sales position will have to lower the price of the product so as to compensate white buyers for their disutility.

5. Statistical Discrimination The concept of taste for discrimination helps us understand how differences between equally skilled blacks and whites (or men and women) can arise in the labor market. It turns out that racial and gender differences will arise even in the absence of prejudice when membership in a particular group (for example, being a black woman) carries information about a person’s skills and productivity.

Statistical discrimination arises because the information gathered from the resume and the interview does not predict perfectly the applicant’s true productivity. The uncertainty encourages the employer to use statistics about the average performance of the group (hence the name statistical discrimination) to predict a particular applicant’s productivity. As a result, applicants from high-productivity groups benefit from their membership in those groups, while applicants from low-productivity groups do not.

6. Measuring Discrimination How economists measure discrimination in the labor market. Suppose that we have two groups of workers, say, male and female. The average male wage is given by , while the average female wage is given by . One possible definition of discrimination is given by the difference in mean wages, or: (8-4)

A more appropriate definition of labor market discrimination compares the wages of equally skilled workers. To simplify the exposition, suppose that only one variable, schooling (which we denote by s), affects earnings. The earnings functions for each of the two groups can then be written as: Male earnings function: Female earnings function: (8-5) The coefficient tell us by how much a man’s earnings increase if he gets 1 more year of schooling, while the coefficient gives the same statistic for a woman. The regression model implies that the raw wage differential can be written as: (8-6)

The Oaxaca Decomposition We can rewrite the raw wage differential as: (8-7) Differential due to discrimination Differential due to difference in skills the second term in the equation arises because the two groups differ in their skills. The first term in the equation arises because of this differential treatment of men and women which is typically defined as discrimination. →The raw wage differential can be decomposed into a portion due to differences in characteristics between the two groups, and a portion that remains unexplained and that we call discrimination.

Figure 3. Measuring the Impact of Discrimination on the Wage Dollars Schooling Men’s Earnings Function Women’s Earnings Function αF αM Figure 3. Measuring the Impact of Discrimination on the Wage