This item is only available as the following downloads:
Gender Discrimination, Economic Segregation and Comparable Worth : Perspectives on the Effectiveness of Policies Targeting the Gender Wage Gap in the United States By Peter Wolfe A Thesis Submitted to the Division of Social Sciences New College of Florida in partial fulfillment of the requirements for the degree Bachelor of Arts in Economics Under the sponsorship of Dr. Patrick Van Horn Sarasota, Florida May 2013
ii Acknowledgements: I wish to thank the friends, family, and professors who have supported my educational career Warf, Varvara Suaraez, Leila Escandar, Lewis Winstanley, Avery Thompson, and Nathan Wilson for bei ng great friends love and support which has been unwavering I would also like to thank Dr. Van Horn, Dr. Coe, and Dr. Khemraj for serving on the committee for this thesis and Dr. Van Horn again for being a supportive thesis sponsor and academic advisor. Finally, I would like to thank Muriel Rice for the years of love and support she gave me, and for instilling an appreciation for Florida in me very early in my life.
iii Table of Contents Title.................................................................................... .............. ..... i Acknowledg ii Table of Contents................................................. ............................. ..iii Abstract .......................................................................... .......... ............ v 1. Introduction.............................................. ....................... ................. 1 The Lilly Ledbetter Act and the Politics of the Gender Wage Gap ..................... ... .. 1 History of the Gender Wage Gap in the United States ........ ........ ............ ..... ...... 3 The Gender Wage Gap Today ........ ........ ............ ............................... ..... 5 2. Literature .10 Human Capital Theory ........ ........ ............ ........................................ ..... 10 Compensating Differentials ........ ........ ............ ............................... ...... .. 14 Discrimination ........ ........ ............ .................. .......................... ... ..... 18 Comparable Worth ........ ........ ............ ............................................ .... 23 3. Comparable W orth: Evidence in National Wage Data The Data ........ ........ ............ ............................... ....... .......... ............ 27 Hypothesi s ........ ........ ............ ............................... ...... ..................... 27 Sex Segregation across Industry Categories ........................................... .... 28 Description of Relevant Variables ........ ........ ............ ............................... 32 Results ....... ........ ............ ............................... ....... ................ ......... 33
iv Analysis ........ ........ ............ ............................... ....... ..... ................. 37 4. Comparable Worth and the Gender Wage Ga p: Policy 39 The Normative Side of Discrimination Policy ........ ........ ............ ................ ... .39 The Republican Perspective ........ ........ ............ ............................... ....... 39 The Democratic Perspective ........ ........ ............ ............................... ..... 43 ........ ........ ............ ............................... ....... ......... 45 Conclusion ....... ........ ............ ............................... ....... ............ ......... 49 Works C
v GENDER DISCRIMINATION, ECONOMIC SEGREGATION, AND COMPATARBLE WORTH Peter Wolfe New College of Florida 2013 ABSTRACT In recent years, t he persistence of the gender wage gap in the United States has been used as a justification to pass legislation strengthening legal protections for victims of gender discrimination in the workplace Opponents of these legislative initiatives argue that our current legal protections against gender discrimination in the wor kplace are strong enough, t hat the gender wage gap in the United States is no longer ca used by discrimination, and that the persistence of the gender wage gap is caused by forces other than discrimination. This thesis takes the position that gender based discrimination is still contributing to expect. Instead of being discriminated against at the individual level, women are discriminated against at the occupational level, wherein o ccupations pay less the more highly female they are, even after controlling for skill requirements and working conditions. Thus, any legislative effort to reduce the gender wage gap ve in reducing the gender wage ga p
1 1. Introduction The Lilly Ledbetter Act and the Politics of the Gender Wage Gap On January 29, 2009, President Obama signed his first bill into law: the Lilly Ledbetter Pay Restoration act. In the presence of reporters, cabinet members, congressional leaders, and Lilly Ledbetter herself, the President made the following remarks: first principles: that we are all created equal and ea ch deserve a chance to pursue for every dollar men earn women of color even less which means that today, in the year 2009, countless women are still losing thousands of dollars in salary, s and their families, we will c lose that pay gap and ensure that our daughters have the same rights, the same chances, and the same freedom to pursue their dreams as our sons (Obama 2009). The Lilly Ledbetter Fair Pay Act was introduced earlier in the month by Senator Clinton and Repre sentative DeLauro, and it was intended to strengthen the legal protections passed in the Equal Pay Act of 1963. Lilly Ledbetter was a Goodyear employee who sued her employer due to unequal payment between her and her male coworkers. The lawsuit Ledbett er v. Goodyear Tire & Rubber Co. eventually reached the Supreme
2 Court, which ruled against Ledbetter on the grounds that Ledbetter did not file suit against her employer within 180 days of when the discrimination began, despite the fact that Ledbetter ha compensation practices long after the 180 day statute of limitations had expired (Ledbetter v. Goodyear, 2007 ). T he Lilly Ledbetter Act was written in response to this ruling, and it changes U.S. law so that the 180 day statute of limitations in discrimination lawsuits begins once the plaintiff becomes aware of discriminatory treatment as opposed to beginning once discriminatory practices first occur. A ddition ally the Lilly Ledbetter Act makes it illegal for employers to prohibit employees from sharing salary information if it is for the purpose of determining if discrimination is occurring ( Perry 155). On January 22, 2009, a week before President Obama signed the Lilly Ledbetter Ac t into law, the Act was up for debate on the floor of the Senate, and the rhetoric was much less positive Republican Senators did not view the Lilly Ledbetter Act in the same positive light as the President On the floor of the Senate, Senator Mitch McC onnell argued that the Lilly Ledbetter Act was not about reducing discrimination at all but was r increase expensive liti gation against business owners, which was unwise since bout these ( Cong. Rec. S739). This sentiment was echoed by other Senate Republicans. However, the Senators who stood in opposition to the Lilly Ledbetter were also persistent in making another point: Senator McConnell: Pay discrimination has been illegal since 1963. Let me say that
3 days on the question of fairness. Every member of this body supports equal pay for equal work. I could not find anybody who does not support that (Cong. Rec. S739). where that is not occurring, let any one of us know, and I bet you we would help to right the wrong. We are against discrimination ( Cong. Rec. S752). History of the Gender Wage Gap in the United States While Senator McConnell and Senator Enzi opposed passage of the Lilly Ledbetter Act, the fact that they felt the need to add the caveat that their opposition to the bill was not based on equal pay for equal work underscores how much attitudes have changed regarding women in the workforce over the past century. In 1940, women earned 59 cents for every dollar earned by a man. Employers believed that the earnings of women were secondary to those of men because wom en were not expected to support families. Throughout most of the history of the United States, paying women less than men was legal. When large numbers of women entered the labor force during World War II, the National War Labor Board suggested that employers equalize wage r ates, but most employers ignored the non binding suggestion (Perry 154). Up until the 1960s, job listings in newspapers were gender specific, with list the same job with different pay rates for men and women. It was not until 1963 with the passage of the Equal Pay Act that it became illegal to pay unequal wages purely on the basis of sex (Perry 153). The Equal Pay Act was an Amendment to the Fair Labor Standards A ct of 1938, work on jobs the performance of
4 which requires equal skill, effort, and responsibility, and which are performed under similar working conditions ( Merrick 21). While the Equal Pay Act was an important step forward for pay equality it was not immediately effective in reducing the gender pay gap. While the bill addressed pay inequity in firms employing both men and women, it did not address wage discrepancies between traditionally male and female occupations (Merrick 23). The broad decl arations of the Equal Pay Act were initially ill defined, further preventing the legislation from effectively combating pay discrimination. A series of court cases in the following decades provided furthe r clarity to the Equal Pay Act and increase d its effectiveness. In 1970, Shultz v. Wheaton Glass Co. came before the United States court of appeals, and the court ruled that equal pay was required for t titles or job descriptions. This ruling prevented employers from hiring women to do the same job as men and justifying lower pay by giving the job a different title. In 1974, the case of Corning Glass Works v. Brennan was heard by the United Stated Sup reme Court, and in their decision the court stated that employers could not pay women a lower wage based on the justification employees. This decision prohibited employers from offering below market rate wages in order to attract female employees (Perry 155). Following these court decisions, the gender dynamics of the labor force began to change dramatically. The change began slowly: in the mid 1970s, women earned about 60 cen ts for every dollar earned by men, which was about equal to the pay gap of the
5 previous 30 years. But by the late 1970s, the gender pay gap began to decrease rapidly, a trend which continued through the 1980s (Burstein, 18). Women were becoming increasin about six percent of MD graduates, and by 1980 that percentage had quadrupled. Similarly, women made up only three percent of JD graduates, whereas today that number has risen to 47% (England 2010 154). By the beginning of the 1990s, the reduction of the gender wage gap began to plateau. From 1993 to 2001, wages for women increased by 0.1 percent per year on average. By comparison, between 1980 and 1993, wages for women increased by 1 percentage point p er year on average (Polachek 23). In the early 1990s, progress on reducing the gender pay gap had come to a halt at about 80 cents earned by women for every male dollar (Blau 1994 23). Very little progress was made in reducing the gender wage gap throu ghout the 1990s and the 2000s, and the gap remains relatively unchanged today (Leicht, 238) (Perry, 154). It seemed as though with the dawn of the 1990s, the gender wage gap had finally converged to a new norm. The Gender Wage Gap Today U.S. Bureau of the Census data from 2003 indicates that the median income for women was $30,599 and for men the median was $40,566. What is especially troubling is the fact that the gap is present even when controlling for levels of education and human capital. In a Congr essional Research Service Report issued by the 109 th Congress, time year round earned $47,910 while time year round earned $69,913 ( Merrick 2).
6 Similarly, female high s chool graduates make $27,956 per year while male high school graduates make $38,331. the central justifications used by supporters of the Lilly Ledbetter Act. The Act itself is seen by its supporters as a small step forward in eliminating the pay gap altogether. And while the opponents of the act also state their support for pay equality, they do not believe that additional action by the federal government should be aimed at reducing the pay gap. While a cynic may believe that Republican opposition to anti discrimination legislation indicates that their rhetoric on gender inequality does not match their actual beliefs, in fact Republicans have a defensible basis for their bel iefs. In January, 2009 the U.S. Department of Labor released a commissioned report prepared pres results of a detailed statistical analysis of the attributes that contribute to the wage gap and a synopsis of the economic research that has been conducted on the ( CONSAD 1). The ultimate conclusion of the exhaustive ninety three page report is the following: differences in the attributes of men and women that account for most of the wage gap (CONSAD this study leads to the unambiguous conclusion that the differences in the compensation of men and women are the result of a multitude of factors and that the raw wage gap should not be used as the basis to justify
7 corrective acti on. Indeed, there may be nothing to correct. The differences in raw wages may be almost entirely the result of the individual choices made by both male and female workers ( CONSAD Economic research has identified many factors that account for portion s of the gender wage gap. Some of the factors are consequences of differences in decisions made by women and men in balancing their work, personal, and family lives. These factors include their human capital development, their work experience, the occupa tions and industries in which they work, and interruptions in their careers ( CONSAD 35). This brings up an interesting possibility, which is that the gender pay gap that remains today may not be due to employer discrimination at all, but rather to societ al attitudes and the division of labor in the home. If this is the case, Republicans may be justified in being wary of burdening employers with legislation that seeks to reduce the gender pay gap. While the CONSAD report is thorough and helps to advance our understanding of the gender wage gap, I believe its assertion that its conclusions are unambiguous is problematic because many of the issues related to the gender wage gap divide the academic community. Despite vast amounts of research on the topic ov er several disciplines, it is not known exactly what causes the gender wage gap, what percentage various factors contributing to gender wage gap account for, and it is not known to what extent employer discrimination plays a role in perpetuating the gap. These are precisely the type of questions which must be answered if the government intends to respond (or not respond) to the gender wage gap effectively. The passage of the Lilly Ledbetter Act certainly seems to imply that for some policymakers, discrimination is still seen as a
8 primary cause of the gender wage gap. Yet the CONSAD Report implies that this view may be misguided based on the results of their research. So who is right and who is wrong? It is my belief that neither side is fully right or fully wrong, but that both tend to overlook key issues that are relevant to the policy debate. More specifically, I think that the problem with both sides of the gender wage gap debate is that they view wage gap has remained stagnant for over two decades now indicates that if we wish to craft policies w hich reduce the gender wage gap in a meaningful way, perhaps we must first redefine the terms of the discussion. Fortunately there is a theoretical approach coming from within both economics and sociology which I believe accomplishes this: the theory of c omparable worth. Unfortunately, comparable worth is generally overlooked by both Democratic and Republican lawmakers and many scholars because it is little understood, and it deviates from how the debate over the gender wage gap is usually framed. The fac t that comparable worth is seldom considered by policymakers is unfortunate considering the fact that the gender wage gap is such an important issue to millions of Americans, as bill signed into law. It remains to be seen if the Lilly Ledbetter Act will serve to reduce the gender wage gap. It empowers women by giving them stronger legal standing in discrimination
9 lawsuits and this may make employers wary of engaging in discrimi natory practices, but the Act also has serious limits. In my view, these limits stem from the fact that the type of discrimination addressed in comparable worth is being overlooked T he lack of action on the part of Republicans and advocated for in the C ONSAD Report overlooks the issues at hand in comparable worth as well. I believe that if we consider comparable worth in our formulation of policy, we could come up with policies that serve to reduce the gender wage gap, establish a more equitable and jus t society, and improve the overall health of the economy. In the remaining chapters of this thesis, my goal is to argue for the use of the theory comparable worth in crafting policies targeting the gender wage gap. In Chapter 2 I review the major theorie s that have been proposed to explain the gender wage gap along with the theory of comparable worth in order to demonstrate how comparable worth contributes to our understanding of the gender wage gap where other leading theories have been less successful. In Chapter 3 I attempt to take the discussion of comparable worth outside of the theoretical realm by showing evidence in national wage data that support the theory of comparable worth, and in Chapter 4 I conclude by discussing how comparable worth ought to influence the policy debate surrounding the gender wage gap in the United States, and discuss additional possible insights comparable worth may have for students of economics and social scientists in general.
10 2. Literature Review S cholars have propo sed a variety of theories in an attempt to explain the persistence of the gender wage gap, and each theory has its proponents and detractors. Some of these theories stand in opposition to each other and some of them overlap, and many scholars believe that various combinations of these theories together explain the entirety of the gap. Which theory or theories one adheres to has major implications for how one would argue that policy makers should address (or not address) the gender wage gap. While there i s variation in how different scholars explain the same theory, broadly speaking there are four main theories that scholars have advocated for in explaining the gender wage gap, and these are human capital theory, the theory of compensating wage differentia ls, the theories of discrimination, and the theory of comparable worth. It is entirely possible that each of these theories explains some portion of the gap; just how much each explains, however, is contested among scholars. For this reason, I will expla in the underlying arguments behind each of these four theories and will examine to what degree each theory has held up under the scrutiny of past scholarly work. This will allow us to question if any of the factors contributing to the gender wage gap migh t be used as a justification for government intervention. Human Capital Theory Human capital theory has been of particular interest to economists studying the gender wage gap. The popularity of human capital theory is understandable: the concept of human capital forms much of the foundation of economic analysis, and thus a model based on human capital would, to an economist, appear to be confirmed by the
11 capacities of hu man beings as income producing agents in the economy Merrick 6). In the neoclassical view of human capital, individuals invest in their level of marketable skills, forgoing present income in exchange for future income. Getting a college education, for instance, costs money and also has an opportunity cost in forgone wages ( England 1992 51). The more one invests in their human capital, the higher their earnings potential (Polachek 3). Human capital theory applied to the gender wage gap posits that the reason the gap exists is that men have higher levels of human ca pital than women. Why might women have lower levels of human capital? Education, one of the primary ways in which individuals increase their human capital, does not explain why a gap in human capital between men and women might arise, because in the U.S. men and women receive, on average, about the same amount of education (England 1992 52), and the median years that men and women spend in school is almost the exact same (Merrick 7). Education, however, is not the only avenue through which workers acqu ire human capital. Human capital is also accumulated through on the job training, and such human capital investments may be firm specific or more general. The cost of training workers in order to increase their productivity is partially paid for by the e mployers, who directly pay for the training, and, as posited by human capital theory, by the workers themselves in the form of lower starting wages (England 1992 51). Thus, if women have, on average, lower levels of human capital than men, then it is lik ely due to a difference in the amount of on the job training that women and men receive.
12 There are a number of reasons why women may receive less on the job training than men, but one of the most studied reasons, particularly among advocates of human capi tal theory, is that women are more likely to take time out of the labor force in order to bear and raise children. This potentially reduces human capital accumulation in a moth erhood may contribute to a gender gap in human capital simply based on the fact that women spend less time in the workplace, and thus have less time to acquire on the job human capital (Blau 2000 6). Going beyond this simple explanation, motherhood may a lso affect the distribution of human capital in other ways. Dropping out of the labor force actually reduces ones incentive to make investments in their own human capital because it makes the opportunity cost of human capital investment higher. Because hu man capital investment necessarily implies a sacrifice on the part of the beneficiary in exchange for some future gain, if an individual decides to drop out of the labor force, that individual is also sacrificing the amount of time in which they are able t o reap the benefits of their human capital investments. In other words, the longer one expects to stay in the labor force, the higher the benefits are of human capital investment. If women are more likely to drop out of the labor force to raise children, rational men and women would choose levels of human capital investment reflecting this. Men, on average, would choose to invest in higher levels of human capital than women, which would increase the productivity of men relative to women. This process wo uld then cause men to command higher wages than women, resulting in a gender wage gap (Polachek, 4) (Blau 2000 6).
13 Thus, the central argument posited by human capital theory is that skill (i.e. human capital) raises ones wage because employers are willin g to pay more for skilled workers because they are more productive, and, on the supply side, employers have to pay higher wages in order to attract workers who have made sacrifices in order to accumulate human capital. Since women are more likely to drop out of the labor force, they accumulate less human capital, and so their wages are lower. Taken in this context, human capital theory provides us with a theoretical framework that implies that a gender gap in wage may exist even in the absence of any disc riminatory behavior on the part of employers. This is important because it implies that legislation attempting to reduce the gender wage gap by penalizing discrimination could prove to be ineffective and potentially burdensome to employers who do not enga ge in discriminatory practices. Thus, the viability of human capital theory has major implications for how policymakers ought to respond to the gender wage gap. Just how well does human capital theory stand up to scholarly scrutiny? Unsurprisingly, the answer to this question is complicated. On one hand, almost all scholars admit that different levels of human capital between men and women contribute to some portion of the gender wage gap. However, estimates as to what percentage of the gender wage gap differences in human capital account for range widely, with some scholars arguing that it explains a relatively small portion the gap, and others, such as Polachek, arguing that it accounts for most, if not all, of the gap. What makes the argument diffic ult to gauge is that human capital proves to be a somewhat allusive variable to measure. There are many different types of human capital, and it is
14 accumulated in many different ways, and as such it does not lend itself to easy measurement (England 1992 69). Thus, out of necessity, much of the debate surrounding human capital theory is of a more theoretical nature than an empirical nature. However, when one considers the theoretical implications of human capital theory, one can begin to tease out ways i n which human capital theory can be tested empirically, and in doing so, serious flaws with the theory as an explanation for the entire gender wage emerge. Compensating Differentials We can better understand the limits of human capital theory by examining another theory used to explain the gender wage gap: compensating differentials. While the theory of compensating differentials is often discussed as a theory separate from human capital theory, in reality the two theories go hand in hand, and the stronges t advocates for human capital theory also tend to advocate for the theory of compensating differentials. The theory of compensating differentials takes the arguments set forth in human capital theory and extends them to include how mothers, as rational ac tors, choose to allocate their labor given the fact that motherhood carries with it a penalty for human capital accumulation. The theory of compensating differentials argues that women expecting to leave the labor force at some point in their lives maximi ze their utility by choosing career paths that are different from those who do not plan on taking such breaks from the labor force. For instance, women who plan on taking a leave from the labor force have an incentive to choose occupations with less on th e job training, since such training will yield lower benefits (Blau 2000 7). Indeed,
15 such choices may be made by women in an effort to maximize their lifetime earnings given the expectation that their employment will be intermittent. Thus, the theory im plies that women have an incentive to choose jobs in which the depreciation of human capital due to time off is relatively low. Human capital theory posits that jobs with high levels of on the job training have lower starting wages than jobs that do not r equire as much training, but that the wage trajectories for these jobs increase more steeply. Therefore women planning on taking time off for child rearing do not benefit from these steep wage trajectories, and therefore they have an incentive to choose j obs with less on the job training, higher starting salaries, and less room for advancement. Women planning to take time out of the labor force will not choose jobs with high levels of on the job training if the gains from steeper wage trajectories do not outweigh the costs of the lower starting wages (England 1992 18). Thus, according to the theory of compensating differentials, the effect of the difference in human capital between men and women gets amplified by segregating males and females into differ ent occupations. A variation of this argument is that mothers or women who expect to be mothers may choose occupations that have lower levels of pay in exchange for nonpecuniary benefits that make their occupations more compatible with their parental resp onsibilities. The more nonpecuniary benefits offered by a job, the lower the wage employers must offer in order to fill positions with qualified employees. The types of nonpecuniary benefits that a parent with child rearing responsibilities may desire in clude flexible work hours, little travel responsibilities, daycare facilities, and safe working conditions (Budig 207). Working mothers with child rearing responsibilities may take these lower wage jobs in
16 exchange for these nonpecuniary benefits while m en would take jobs with less nonpecuniary benefits but higher wages. If this is the case, it offers another reason why a gender wage gap could exist in the absence of employer discrimination. Unlike the more broad assertions posited by human capital theo ry, the theory of compensating differentials is open to empirical testing, and numerous scholars have tested the validity of the theory of compensating differentials. Most of this work has not supported the notion that the theory of compensating differentials explains a significant portion of the gender wage gap as it exists today. On the surface, the theory of compensating differentials seems promising in explaining certain trends in wage dat a. For instance, it has been found that on average, mothers make less money than non mothers even when one controls for factors such as part time status, seniority, and experience. However, digging deeper into the data, scholars find that the occupations women hold do not, on average, have the supposed nonpecuniary benefits that would make these jobs more compatible with motherhood (England 2005 277). In fact, data seem to indicate that the opposite may be true, as men are more likely than women to have flexible hours, and in addition men tend to have both more unsupervised break time, more days of paid sick leave, and more vacation days than women the very factors one would assume that a parent would desire if they were trying to balance work responsi bilities and family responsibilities (England 1992 14). Analyzing this data from the opposite direction that is, seeing if nonpecuniary family friendly fringe benefits have a negative effect on wages also calls into question the validity of compensat ing differentials. Data shows that jobs that include nonpecuniary fringe
17 benefits tend not to have a wage penalty for the receipt of these benefits. Fringe benefits are most correlated with positions of authority, and thus we find that men tend receive t hese benefits more often than women, since men are more likely to hold positions of authority ( Merrick 12). Finally, data has also shown that non mothers, who, as mentioned previously tend to make more than mothers are just as likely to be concentrated into traditionally female jobs as are mothers, which calls into question the notion that mothers choose such female jobs due to their status as mothe rs (England 2005 277). What of the related argument that women choose jobs with higher starting wages to avoid losses resulting in the depreciation of human capital? If this is the case, then women planning on taking time out of the labor force may simply be attempting to maxim ize their lifetime earnings by entering jobs with less training and higher starting wages, but flatter wage trajectories. However, scholarly research addressing this argument has not supported it. Scholars have found that in predominantly female occupati ons, starting wages are in fact lower than they are in predominantly male occupations, and in fact some research suggests that women are more likely to receive on the job training than men (England 2005 271). Therefore, the choices that human capital the orists predict women will make that lead to occupational segregation between men and women do not hold up to scrutiny, as the motivations that human capital theorists predict would lead women to choose different jobs from men do not actually exist, and thi s is true net of work hours, work interruptions, part time work, and parenthood status (England 1992, 53) ( Merrick 8).
18 Discrimination If human capital theory and wage differentials cannot explain the root causes of the gender wage gap, then what can? Th eories stemming from the human capital model tend to be juxtaposed with theories that point to discrimination as the root cause of the gender wage gap. It is important to note, however, that these theories are not necessarily mutually exclusive, and that both human capital theory and theories based on discrimination may both be at work in contribution to a portion of the gender wage gap (England 1992 52). Importantly, the portion of the gap that these theories explain may be overlapping, as a small amoun t of gender discrimination in the workplace may influence the division of labor in the home. If women are more likely to be discriminated against in the workplace, this could reinforce the trend of women dropping out of the labor force to take on parentho od duties instead of men as households attempt to maximize their total income. In other words, discrimination could lead to women accruing less human capital than men because the opportunity cost for women dropping out of the labor force is lower than it is for men, because women would face gender discrimination and men would not. Then furthermore, this may make employer discrimination more likely women are more likely to drop out of the labor force (Albanesi 81). An important aspect of this feedback effect is that is gives rise to the possibility that even past discrimination could be contributing to the gender wage gap today by leading women to hold jobs with less seniority. To better understand the role that gender discrimination plays in the gender wage gap, it is important that we understand how discrimination is
19 viewed in economic theory, and to see what research, if any, suggests that discrimination contributes to the gender wage gap (England 1992 54). Discrim ination can come in a variety of forms. Under U.S. federal law, two forms of discrimination are prohibited. The first type of discrimination that is prohibited is called differential treatment, which occurs when employers treat female employees different ly than male employees because of their sex as opposed to their qualifications, either in hiring or in promotions. The second form of discrimination that is illegal under U.S. law is discrimination that has a disparate impact on women. This type of discr imination would mean that employers screen applicants for jobs and/or promotions based on criteria other than sex, but that tends to favor one sex largely over the other, provided that the criteria being used by the employer is not based on a that is, any criteria that serves to increase productivity and/or reduce business expenses (Budig 209). Of course, the fact that these discriminatory practices are illegal does not mean that they do not occur, and by extension their prohibition gap. Within the field of economics, among the first to theorize about the role of discrimination in competitive labor markets was Gary Becker, who developed a model of discriminat ion in which discrimination is caused by discriminatory tastes on the part of employers, customers, or co workers (Blau 2000 7). In neo classical economic theory, one has more of a taste for something if one is willing to pay extra money for it. Therefor e, taste discrimination implies that one is willing to pay more to hire people
20 from a certain group than they are from another group, even if the members of each exogenous v ariable that is formed by premarket factors. If an employer has a taste for discrimination against women, then they will opt not to hire any women unless the wage paid to women is lower than the wage paid to men. The size of the discount that an employer demands depends upon the strength of their preferences. At one extreme, some employers could have a taste for discrimination that leads them to higher no female labor, even if female labor is free. At the other extreme, someone with no taste for discrim ination will hire equally qualified men and women at the same wage. In this view, hiring discrimination and wage discrimination are both outcomes resulting from the same type of discrimination: taste discrimination (England 1992 54). discrimination will erode from competitive markets, since firms run by employers with non discriminatory tastes should be able to hire women and take advantage of the lower wages required for their labor due to crowding in the female labor market (Blau 2000 11). Employers with the most taste for discrimination will hire only men, while employers with some taste for discrimination will hire women if the difference the wages paid to men an d women is large enough to offset their tastes, and firms with no taste for discrimination will hire only women at a lower wage. Thus, when we use the to take advanta ge of the exploited status of women for economic gain (England 1992 62). Because of the lower labor costs paid by non discriminatory employers, they are
21 given a competitive advantage over firms with tastes for discrimination because non discriminatory fi therefore predicts that in the long run, firms with a taste for discrimination will be outcompeted by non discriminators until eventually only non discriminators are left, and they prediction that discrimination sows the seeds of its own destruction has led some economists to deemphasize the role that discrimination plays in explaining the gender wage gap ( England 2005 (England 1992 63). Economists have been fairly unsuccessful in determining the extent of gender based hiring and placement discrimination. Part of the reason for this is that much of the data coming from large survey samples only show what jobs women hold, not the jobs they have applied for (England 2005 272). get empirical evidence on discrimination, p articularly since it is illegal in the United States to act on discriminatory tastes if you are an employer. Scholars have offered alternative models of discrimination to the Becker model, and the model advocated for by many researchers is the model of st atistical discrimination, which states that, because of the intermittent employment of mothers, women are discriminated against regardless of parenthood status because on average women are more likely to have parental obligations than men (Merrick 13). T his model relies on the assumption it is costly to screen applicants before hiring, so in the absence of perfect information, they discrimi nate against women because they may be or become mothers (Budig 209).
22 tatistical discrimination will not necessarily erode in the long run because it may be profit maximizing for employers to engage in statistical discrimination. Employers may find it cheaper and easier to hire one gender over the other rather than screen t he exact qualifications of each individual applicant, particularly in jobs in which men are more likely to have the proper educational credentials than women (England 2005 273). Scholars have been able to gather little empirical data supporting the model of statistical discrimination. First, it is unclear that women actually have higher turnover rates than men. While women are more likely than men to leave a job because of parental responsibilities, men are more likely than women to leave their jobs in order to serve in the military, serve a prison sentence, and men are also more likely to change firms. Thus, there is little incentive for employers to discriminate against women because of a higher turnover rate when a higher turnover rate does not actua lly exist (England 2005 274). Experimental data has found that some statistical discrimination does indeed still occur. In studies where identically described men and women apply for the same job, it was found that men were much more likely to be hired than women. The problem is, these studies show discrimination only within specific occupations, so more of these studies are needed if we are to know the extent of statistical discrimination in the labor market (England 2005 272). However, overall the e vidence suggests that while statistical discrimination still occurs, it is likely that other factors play a larger role in causing the gender wage gap (Merrick 15).
23 Comparable Worth So far we have examined human capital theory, compensating differential s, and each of these theories. The conclusion we come to regarding these theories is that while they all may be causing a portion of the gender wage gap, none of them are able to explain the primary cause of the gap. We have seen different ways in which these theories overlap and ways in which they oppose each other, but their ultimate inability to explain the gender wage gap begs the need for a more dynamic theory that c an explain the gender wage gap more fully and that can be supported by empirical evidence. There is one commonality linking each theory so far discussed: that women and men tend to be employed in different occupations. The explanations for why occupatio nal segregation by gender occurs differ depending on which theory one adheres to, but the fact that is does indeed occur is found consistently across each theory. Empirical data fully supports this, showing that men are more highly represented in upper ma nagement positions, blue collar craft jobs, heavy manufacturing, transportation, and construction. Women, on the other hand, are heavily represented in the fields of nursing, teaching (k 12), librarianship, clerical work, non commission retail sales, manu facturing in nondurable goods, domestic work, and childcare (England 2005 266). Analysis of wages between male and female dominated jobs reveals that the jobs dominated by women have lower wages than jobs that are dominated by men (Perry 156).
24 The theor y of comparable worth provides us with insights into how gender discrimination, occupational segregation, and the gender wage gap are interrelated. Stated simply, the theory of comparable worth posits that jobs filled mostly by women pay lower wages than they would if those jobs were filled mostly by men. Jobs that are traditionally female are paid less than jobs that are traditionally male, net of required educational levels, skill demands, and working conditions. Thus, comparable worth argues that ther e is a form of gender based discrimination that occurs when certain occupations those dominated by females have their wages set (England 1992 1). On the surface, comparable worth sounds similar to theories of discrimination where women are given unequal pay for equal work; where equally qualified men and women holding the same job are paid differently. Comparable worth is distinct from this issue in that compares the wages of men and women in different occupations that is, it compares wages in traditionally female jobs to wages in traditionally male jobs. In making this comparison, scholars have found that female jobs pay lower than male jobs n et of skill demands and working conditions. Comparable worth does argue that gender discrimination is to blame for the difference in pay between men and women, but it is a different type of discrimination than the discrimination in the theories of discrim ination discussed so far. Discrimination in the theory of comparable worth refers to a gender bias in setting wages, as opposed to a gender bias in hiring, initial job placement, or promotion (England 1992 1). Researchers have found when testing for the degree of segregation by sex, the overall level of segregation increases as one uses more detailed occupational categories. If one measures the level of segregation based on what
25 industry they work in, then the level of occupational segregation will seem lower than if one separates out different occupational categories within industries (England 1992 13). Additionally, the higher percentage of females within an occupation, generally speaking the lower the wages are in that occupation for both men and wo men (England 1992 125). Why is it important to distinguish between the gender discrimination proposed by the theory of comparable worth and other forms of gender discrimination? It is important because discrimination in hiring, placement and promotion al l would be considered a violation of the Civil Rights act, whereas the type of discrimination that occurs in the theory of comparable worth is perfectly legal. The pro male and/or anti female gender biases of employers that lead employers to set lower wag es for female dominated jobs is legal because men who occupy such female dominated jobs will suffer the same wage penalty as wom en in those jobs (England 2005 276). Both cultural and institutional forces contribute to employers devaluing the work done by women and underestimating the role that female jobs play in organizational goals, giving these jobs a lower wage. And, once the lower market wages for female jobs have been established, the devaluation becomes self perpetuating as firms use past wages as the basis for present wages (England 2005 277). Comparable worth theory holds a special significance for how we address the CONSAD Corporations claims that the existence of a wage gap should not be used as a justification for new legislative action in a way that the other theories explaining the gender wage gap do not. I do not put forth the argument that comparable worth
26 explains all or even necessarily most of the gender wage gap. However, if it explains any portion of the gender wage gap, then it imp lies a need for new legislation in a way that none of the other theories do. Human capital theory and the theory of compensating differentials imply that the existence of the gender wage gap is the result of choices made by men and women, which would impl y that the gender wage gap is not a of statistical discrimination attribute the existence of the gender wage gap to discriminatory practices that are already illegal an d have been for several decades, implying that all that is needed to reduce the gender wage gap is better enforcement of existing laws. The discrimination at issue in comparable worth is both discriminatory and legal, and thus stands as the one theory on the gender wage gap that might warrant new legislative initiatives.
27 3. Comparable Worth: Evidence in National Wage Data The Data The data used in this chapter comes from the Center for Economic and Policy Research (CEPR) Current Population Survey for the year 2010. The Current Population Survey is used by the U.S. Government to track unemployment and labor force participation. The extensive amount of information and different variables measured by the Current Population Survey provides us with a wealth of relevant information for analyzing the gender wage gap Hypothesis I believe that this data will support the general hypothesis posited by the theory of comparable worth that is, that sex has a negative effect on wages of female occupa tions even when other variables affecting wage are controlled for. As noted in Chapter 2, past studies have consistently found that occupations are segregated by sex, and that female jobs tend to pay less than male jobs net of skill requirements and worki ng conditions. These studies tend to focus on occupations as opposed to industries, since within industries, one finds a variety of occupations that may or may not be gendered (England 1992 13). While this approach has yielded useful results that seem t o be consistent with the theory of comparable worth, the focus on occupation as opposed to industry has also left certain gaps in our understanding of how comparable worth policies might be successfully applied in a complex modern economy. Specifically, t he focus of comparable worth research on occupation as opposed to industry has neglected to address the possibility that within industries, a specific occupation may or may not be gendered. Comparable worth policies that fail to
28 take this consideration in to account risk being either too broad or not broad enough in The intent of this chapter is to incorporate the effect that occupational segregation has on the earnings o f male and female workers across industries, and in doing so it is my beyond aggregate sex composition. I predict that (1) industries are segregated by sex similarly to o ccupations, (2) that females will tend to hold traditionally female occupations across industries and men will hold traditionally male occupations, regardless of the level of gender segregation within an industry, and therefore that (4) be relatively unaffected by the industry they work in, whereas work in. This, I argue, will provide evidence that occupations are segregated similarly across indus tries; that female occupations are generally devalued compared to male occupations even after controlling for human capital requirements, nonpecuniary compensation, and working conditions, and that comparable worth policies can be effective across industry categories. Sex Segregation across Industry Categories Figure 3.1 gives us a good visual representation of which industries are most segregated by sex. Table 3.1 gives a detailed breakdown of these results. As we can see, industries are segregated by s ex similarly to occupations. Seven industries stand out as being over 65% female. These are, in order from most to least segregated, Private Households (91.54% female), Social Assistance (86.50% female), Health Care Services, except Hospitals (79.29% fem ale), Hospitals (77.57% female), Other
29 Information Services (76.35% female), Personal and Laundry Services (74.45% female), and Educational Services (68.87% female). Conversely, there Figure 3.1 Table 3.1 Industry Industr y Code Total Female Male % Total Female Agriculture 1 1,722 439 1,283 25.49% Forest ry, logging, fishing, hunting 2 166 34 132 20.48% Mining 3 786 83 703 10.56% Construction 4 6,712 621 6,091 9.25% Nonmetallic mineral product m fg 5 292 60 232 20.55% Primary metals and fabricated metal pro 6 1,083 184 899 16.99% Machinery manufacturing 7 821 178 643 21.68% Computer and electronic product m fg. 8 840 286 554 34.05% Electrical equipment, appliance m fg 9 277 99 178 35.74%
30 Transportation equipment m fg 10 1,305 303 1,002 23.22% Wood products 11 273 39 234 14.29% Furniture and fixtures m fg 12 292 85 207 29.11% Miscellaneous and not specified m fg 13 993 402 591 40.48% Food manufacturing 14 1,401 559 842 39.90% Beverage and tobacco products 15 158 43 115 27.22% Textile, apparel, and leather mfg 16 396 224 172 56.57% Paper and printing 17 620 197 423 31.77% Petroleum and coal products m fg 18 141 25 116 17.73% Chemical manufacturing 19 806 286 520 35.48% Plastics and rubber products 20 312 98 214 31.41% Wholesale trade 21 2,384 692 1,692 29.03% Retail trade 22 11,169 5,590 5,579 50.05% Transportation and warehousing 23 3,972 1,003 2,969 25.25% Utilities 24 803 177 626 22.04% Publishing industries (except internet) 25 408 203 205 49.75% Motion picture and sound recording ind 26 247 83 164 33.60% Broadcasting (except internet) 27 408 143 265 35.05% Internet publishing and broadcasting 28 41 12 29 29.27% Telecommunications 29 718 242 476 33.70% Internet service providers and data pro 30 76 38 38 50.00% Other information services 31 148 113 35 76.35% Finance 32 2,997 1,678 1,319 55.99% Insurance 33 1,670 996 674 59.64% Real estate 34 1,528 754 774 49.35% Rental and leasing services 35 258 69 189 26.74% Professional and technical services 36 6,380 2,828 3,552 44.33% Mgt of companies and enterprises 37 122 64 58 52.46% Administrative and support services 38 4,241 1,783 2,458 42.04% Waste mgt and remediation servic es 39 339 60 279 17.70% Educational services 40 9,114 6,277 2,837 68.87% Hospitals 41 4,089 3,172 917 77.57% Health care services, except hospitals 42 6,933 5,497 1,436 79.29% Social assistance 43 2,437 2,108 329 86.50% Arts, entertainment, and recreation 44 2,079 979 1,100 47.09% Accommodation 45 1,172 693 479 59.13% Food services and drinking places 46 6,086 3,221 2,865 52.92% Repair and maintenance 47 1,410 162 1,248 11.49% Personal and laundry services 48 1,695 1,262 433 74.45% Membership associations and organizatio ns 49 1,351 770 581 56.99% Private households 50 508 465 43 91.54% Public administration 51 4,765 2,237 2,528 46.95% Armed forces 52 27 5 22 18.52% are seven industries in which women make up less than 20% of the total workforce. These are, from most to least segregated, Construction (9.25% female), Repair and
31 Maintenance (11.49% female), Wood Products (14.29% female), Primary Metal and Fabricated Metal Products (16.99% female), Waste Management and Remediation Services (17.70% female), Petroleum and Coal Products Manufacturing (17.73% female), and Armed Forces (18.52% female). This gender distribution across industry categories is consistent with the f indings reviewed in Chapter 2 regarding the gender distribution across occupational categories that is, that women tend to be most heavily represented in the fields of nursing, teaching (k 12), librarianship, clerical work, domestic work, and childcare, whereas men are more highly represented in blue collar craft jobs, heavy manufacturing, transportation, and construction (black, 266). So as we see, while many industries are integrated, there are many others in which there is still a high degree of segre gation by gender, and they tend to follow the same trends of segregation found when we look at segregation by occupation. When looking at gender segregation by industry, we cannot say if male or female industries have higher or lower skill requirements, since within each industry a range of different occupations are represented that span all levels of skill requirements. Thus, for both men and women, we would expect that after controlling for human capital characteristics and compensating differentials, to the profitability of the industry in which they work. However, if comparable worth is correct, and the occupations held by primarily by women within each industry are valued less than the occupations held prima rily by men, then even after controlling for these characteristics we would expect industry to have a much larger effect on the wages of males than on the wages of females.
32 In order to test this hypothesis, I conducted two OLS Regressions in order to see t he effect that working in a specific industry has upon wages while controlling for human capital characteristics, nonpecuniary compensation, and working conditions. The two Regressions are identical except that one is run for female respondents and the ot her for male respondents. Description of Relevant Variables B efore we are able to delve into the data, I will first explain the relevant variables for our analysis. Since we are testing how industry affects wage it is natural that ou r dependent variabl e be wage For our purposes, we will measure wage based on the natural log wage rate, so that the effect that the independent variables have on wage will be shown in percentage terms For the independent explanatory variables, we must first consider which variables will affect wage other than industry. H uman capital theory must consider which variables best measure human capital. T he easies t measure is educational attainment. Human capital theory argues that human capital is also acquired through work experience. Since there is not a variable that measures work Advocates of human capital theory argue that a major reason women attain less work experience than men is because women are more likely to drop out of the labor force to raise children. Thus, by including parenthood status as an independent variable, we can control for reductions i n work experience resulting from parenthood. Additionally,
33 marriage status is controlled for, since married parents are likely able to take off from work for longer periods than single parents. The theory of compensating differentials argues that women, p articularly mothers or women who plan on becoming mothers, may take jobs with lower wages but more stability in order to have careers that are more compatible with the responsibilities of motherhood. What factors make a job more compatible with family res ponsibilities is somewhat subjective, but there are certain factors we can assume working parents would value. First, compensating differentials argues that public sector jobs provide more stability and flexibility than private sector jobs, and as a resul t have lower wages but more amicable working conditions. Thus, public/private sector employment status is another variable that must be controlled for. Since parents may n provides paid leave. If compensating differentials is correct, then the availability of paid leave as a nonpecuniary benefit should come at a reduction in wages. Union membership is also controlled for, as we would expect union members to have higher w ages than nonunion members. Finally, the regression includes dummy variables for each industry listed in our data. Thus, formally our regression states: ln_wage = c + age +marstat + child + unmem + pubsect + wbho + rincp_oth + industry Results The result s of our regression for females are summarized by table 3.2 and for males by table 3.3. What do our results indicate? First, for both males and females, age, union membership, and education are all shown to be correlated with an increase in wages, as we would expect under human capital theory. However, surprisingly, while having
34 children impacts wages for men and women both negatively that impact is shown to be significant for males, yet insignificant for females. This indicates that the importance tha t human capital theory places on the role of motherhood in perpetuating the gender wage gap is likely overstated. Table 3.2 (MALES) *indicates significance at 1%; **indicates significance at 5% Dependent Variable Coefficient t statistic Age 0.007 (11.32)** Race 0.04 (5.23)** Marital Status 0.056 (11.99)** Children 0.293 (3.71)** Union Membership 0.171 (7.34)** Public Sector Employment 0.013 0.43 Years of Education 0.23 (30.93)** Income from other Sources 0 0.08 Forestry, logging, fishing, hunting, an 0.382 1.85 Mining 0.696 (8.29)** Construction 0.52 (8.16)** Nonmetallic mineral product manufacturi 0.416 (3.78)** Primary metals and fabricated metal pro 0.506 (6.66)** Machinery manufacturing 0.496 (5.68)** Computer and electronic product manufac 0.644 (7.50)** Electrical equipment, appliance manufac 0.571 (4.73)** Transportation equipment manufacturing 0.569 (7.42)** Wood products 0.534 (4.21)** Furniture and fixtures manufacturing 0.365 (2.99)** Miscellaneous and not specified manufac 0.479 (5.25)** Food manufacturing 0.292 (3.48)** Beverage and tobacco products 0.489 (3.22)** Textile, apparel, and leather manufactu 0.283 (2.17)* Paper and printing 0.452 (5.11)** Petroleum and coal products manufacturi 0.584 (3.22)** Chemical manufacturing 0.617 (7.07)** Plastics and rubber products 0.378 (3.24)** Wholesale trade 0.437 (6.20)** Retail trade 0.248 (3.95)** Transportation and warehousing 0.393 (5.88)** Utilities 0.604 (7.38)** Publishing industries (except internet) 0.298 (2.35)* Motion picture and sound recording indu 0.458 (2.53)* Broadcasting (except internet) 0.557 (5.05)** Internet publishing and broadcasting 0 (.) Telecommunications 0.67 (7.67)**
35 Internet service providers and data pro 0.466 (2.01)* Other information services 0.256 0.94 Finance 0.715 (9.76)** Insurance 0.464 (5.44)** Real estate 0.408 (4.56)** Rental and leasing services 0.249 1.78 Professional and technical services 0.627 (9.43)** Management of companies and enterprises 0.618 (2.48)* Administrative and support services 0.319 (4.56)** Waste management and remediation servic 0.347 (3.18)** Educational services 0.148 (2.14)* Hospitals 0.434 (5.53)** Health care services, except hospitals 0.402 (5.42)** Social assistance 0.05 0.48 Arts, entertainment, and recreation 0.306 (3.91)** Accommodation 0.177 1.82 Food services and drinking places 0.117 1.75 Repair and maintenance 0.418 (5.36)** Personal and laundry services 0.325 (3.03)** Membership associations and organizatio 0.155 1.75 Private households 0.197 0.85 Public administration 0.502 (6.98)** Armed forces 0 (.) Constant 1.789 (25.51)** Observations: 6693; R squared: 0.34 Table 3.3 (FEMALES) *indicates significance at 1%; **indicates significance at 5% Dependent Variable Coefficient t statistic Age 0.006 (10.38)** Race 0.006 0.82 Marital Status 0.031 (6.54)** Children 0.069 0.94 Union Membership 0.125 (4.86)** Public Sector Employment 0.017 0.65 Years of Education 0.242 (32.31)** Income from other Sources 0 (2.78)** Forestry, logging, fishing, hunting, an 0 (.) Mining 0.268 1.2 Construction 0.121 0.87 Nonmetallic mineral product manufacturi 0.113 0.51 Primary metals and fabricated metal pro 0.071 0.43 Machinery manufacturing 0.37 (2.13)* Computer and electronic product manufac 0.343 (2.17)* Electrical equipment, appliance manufac 0.265 1.3 Transportation equipment manufacturing 0.376 (2.53)* Wood products 0.392 1.22 Furniture and fixtures manufacturing 0.105 0.47
36 Miscellaneous and not specified manufac 0.214 1.5 Food manufacturing 0.059 0.41 Beverage and tobacco products 0.311 1.15 Textile, apparel, and leather manufactu 0.081 0.5 Paper and printing 0.164 1.02 Petroleum and coal products manufacturi 0 (.) Chemical manufacturing 0.46 (3.04)** Plastics and rubber products 0.251 1.08 Wholesale trade 0.169 1.29 Retail trade 0.027 0.22 Transportation and warehousing 0.237 1.83 Utilities 0.259 1.63 Publishing industries (except internet) 0.267 1.58 Motion picture and sound recording indu 0.297 1.1 Broadcasting (except internet) 0.513 (2.98)** Internet publishing and broadcasting 0 (.) Telecommunications 0.263 1.68 Internet service providers and data pro 0.24 0.82 Other information services 0.011 0.06 Finance 0.285 (2.29)* Insurance 0.185 1.43 Real estate 0.219 1.63 Rental and leasing services 0.597 (2.05)* Professional and technical services 0.345 (2.80)** Management of companies and enterprises 0.316 1.24 Administrative and support services 0.044 0.35 Waste management and remediation servic 0.404 1.59 Educational services 0.05 0.41 Hospitals 0.346 (2.83)** Health care services, except hospitals 0.094 0.78 Social assistance 0.09 0.72 Arts, entertainment, and recreation 0.042 0.31 Accommodation 0.132 0.98 Food services and drinking places 0.111 0.9 Repair and maintenance 0.35 1.9 Personal and laundry services 0.149 1.15 Membership associations and organizatio 0.046 0.35 Private households 0.084 0.56 Public administration 0.241 1.92 Armed forces 0 (.) Constant 1.706 (13.48)** Observations: 6629; R squared: 0.29 What to the results of our analysis indicate about the effect of industry upon much more significant impact on the
37 wages for men than it does for women. Of the 52 industry categories included in our regre ssion, female wages are shown to be significantly impacted by employment in 5 industries at 1% significance, and in 4 industries at 5% significance. By comparison, indust ries at 5% significance. It is clear that for men, industry has a substantially larger impact on wages than it does for women. This is consistent with what comparable worth would predict although it worth noting that the effects of parenthood and race i n particular were found to be surprisingly low. These results indicate that there is likely a high degree of occupational segregation by gender within industries, and that for men, being in a highly profitable industry tends to lead to much higher wages, whereas for women, which industry one is employed in is more or less irrelevant to wage. This productivity. This is consistent with the underlying notion of comparable worth that the goals of the employers for which they work. Analysis These results are hardly surpris ing considering which occupations women are most likely to hold. It seems like there is a general devaluation of work that involves care, nurturing, communication, and organization. Perhaps the reason these skills are not highly valued is because it is h ard for the employee who exercises such skills to control the full value created by their labor. Take teachers, for instance though teaching k
38 educated population is something that be nefits everyone in society, but not everyone who benefits from the work done by teachers pays for the full benefits they reap, and teachers are paid a lower wage than workers in male occupations who are similarly educated. This indicates that the lower wa ges paid to female jobs may be representative of what economists refer to as the free rider problem. Taken further, this line of reason may shed light on the underlying social causes that lead to the devaluation of female work more broadly. Take motherho od, for instance. Everyone in society benefits from good parenting through reduced crime and a more productive population, yet mothers who take time off of the labor force to raise children bear the full costs of parenting through both material expenditures and through lost earning s. Perhaps it is unwise for a society to penalize work that has positive social impacts as from an economic perspective and from a social justice perspective, and comparabl e worth may be the way we can correct this incentive problem.
39 4. Comparable Worth and the Gender Wage Gap: Policy Implications The Normative Side of Discrimination Policy Does comparable worth constitute a form of d iscrimination that justifies legislativ normative analysis. How one feels about the issues at hand in comparable worth and the gender wage gap more generally hinge upon moral and ethical judgments regarding the appropriate role of the government, gender dynamics, and economic justice. Once these normative values are established, positive analysis can be used to determine what the policy i ssues referred to in Chapter 1 regarding the Lily Ledbetter Act and the CONSAD report. In the statement given by President Obama in support of the Lily Ledbetter Act, the statements made by Congressional Republicans in opposition to it, and in the conclus ion put forward in the CONSAD report, we see different normative values being expressed that lead to different ideas about how the government should respond to the gender wage gap. In this chapter, I will discuss how these different normative beliefs lead to different opinions regarding the proper policy response to the on the policy stances one takes. I will then conclude by explaining how the theory of comparable worth has impacted my own understanding of the gender wage gap and how I believe policymakers ought to respond to the gender wage gap. The Republican Perspective Let us begin this discussion by referring back to the statements made by Congressional Republicans in opposition to the Lily Ledbetter Act. In the comments
40 made by both Senator McConnell and Senator Enzi quoted in Chapter 1, both Senators indicate that they support government intervention in the economy for the purposes of preventing wage discriminatio n. Their purpose in emphasizing this point, I believe, is to clarify that their opposition to the Lily Ledbetter Act is not due to a belief in strict libertarianism and/or complete laissez faire the economy. Both notions are rejected in their expression of support for Equal Pay Act of 1963, which made gender based hiring and placement discrimination illegal in the United States. Both libertarianism a nd laissez faire economics, if followed strictly, would imply opposition to the Equal Pay Act of 1963 for similar, although not identical, reasons. In the case of libertarianism, one would reject the Equal Pay Act on the basis of what constitutes a person al right libertarianism believes that the government should play a role in protecting individual rights, but in a very limited capacity: it should enforce contracts and protect its citizens from violence, but that any government control over individual a ction beyond this constitutes a violation of individual liberty. From this perspective, the Equal Pay Act constitutes a violation of individual rights in that it prevents employers from engaging in wage discrimination in their own businesses, which they h ave a right to do since they have a right to have a preference for discrimination. Laissez faire economists, on the other hand, believe that unregulated markets will lead to the most efficient outcomes as rational actors make mutually utility maximizing e xchanges, and therefore that government intervention in the economy will lead to less efficient market outcomes in the long autonomy in hiring decisions as is done in the Civil Rights Act, therefore, would be
41 thought to prevent the economy from reaching its maximally efficient long run equilibrium. Most Republicans and Democrats today hold the position that while free exchange does lead to the most efficient economic outcomes, there are some policy goals such as preve nting discrimination that do justify government intervention into the economy. There are a number of different reasons why one might reject a strict laissez faire or libertarian view and support anti discrimination legislation. For instance, one might believe that the underlying assumptions in the laissez faire approach are inapplicable to certain situations, and thus the market could lead to inefficient outcomes in the absence of government intervention. One might also disagree with the libertarian vi ewpoint of individual rights being primarily based on freedom from government, and instead take a more egalitarian of view of individual rights and believe that individuals have a right to the same economic opportunities as everyone else. Senators McConne ll and Enzi both express support for anti discrimination legislation based on the concept of the egalitarian notion of equality of opportunity in mentioning increases legal p rotections against gender discrimination, and their explanation for why they opposed the Lily Ledbetter Act primarily cites the potential economic impact of the legislation: Senator McConnell: It unfairly targets business owners who, in many cases, will no longer have the evidence they will need to mount a just defense. As we all know, job creators have enough to worry about these days. We should not add the t hreat of never If we are going to grow our economy, we need to focus on
42 legis lation that will create jobs, not put undue hardships on job creators ( Cong. Rec., S739). Senator Inhofe: What we are told by the other side of the aisle is that the Lilly Ledbetter Fair Pay Act is about protecting the right of employees who may not know they have been discriminated against. But in reality, this bill represents a tremendous burden on employers and a boon for t Tomorrow in Oklahoma I will be meeting with voters in Clinton and Burns Flat and other areas in sou thern Oklahoma. It will be my unfortunate duty to tell them that this burden has been unfairly placed upon them and their businesses in this difficult economic time ( Cong. Rec., S748). Senator Enzi: What we are looking at here could be an exponential incre ase in lawsuits at a time when many employers are struggling to make their payroll and avoid laying people off A small businessman faced with a lawsuit that is going to cost him $20,000, $25,000, $100,000 to defend cannot afford the time or the money to d o that and may work harder at a settlement and encourage people to do lawsuits that may not have the same merit we are trying to achieve in this bill ( Cong. Rec., We are against discrimination. But we are also against discrimination against the small businessmen who have to sometimes interpret our laws, figure out what we are saying, and become some of the precedent setters on some of the fine them. What they are trying to do is put out a product or service and get compensated for it so they can compensate their employees. There are a lot of decisions they have to make to be able to d o that. Fairness is one of them ( Cong. Rec., S752). As these quotes demonstrate, R epublican opposition to the Lily Ledbetter Act relies heavily on the justification that it would have a negative impact on the economy by
43 increasing the expenses businesses would have to pay in legal fees. What is seldom addressed in their statements is w hether the Lily Ledbetter Act actually would reduce important caveat that they are leaving out, which is that they support equal pay as long come at too high where this line is drawn. Additionally, the statements from Congressional Republicans in Chapter 1 imply that they believe that the Lily Ledbetter Act is not needed because U.S. law already makes hi ring and placement discrimination illegal. However, since there is still a wage gap between men and women, what exactly is meant by their use of the there should be no gen der gap at all, or does it imply only that equally productive men and women with the same job in the same business be paid the equally? Based solely exactly what the Republican position is reg ought to be in ensuring it, and because of this we are unable to know under what circumstances Republicans would consider adopting comparable worth policies. But it does show that, a t least for certain Republicans, the existence of a gender wage gap does not justify legislation meant to reduce the gap that would be, in their view, overly cumbersome to employers. The Democratic Perspective The Lilly Ledbetter Act is about a technicalit y over the statute of limitations in employment discrimination lawsuits, and thus ones opposition or support for the law
44 comparable worth policies. This makes sense, since the Li supporters, the existence of the gender wage gap is the primary justification for the law. atements are generally just about correcti ng a judicial statute that is gap is itself unfair. While the President acknowledges the Lilly Ledbetter Act alone will not eliminate the gender wage gap, it is importan t nonetheless because it is a step toward eliminating the gender wage gap altogether, in his view. As with the arguments against for the Lily Ledbetter Act reveal a number of normati ve assumptions about what discrimination is Congressional Republicans, President Obama expresses a view of individual rights that are egalitarian, and the government has a role in regulating the economy in order to close that pay gap and ensure that our daughters have the same rights, the same chances, and the same freedom to pursue their dreams as our sons Lilly Ledbetter Act is a step toward closing the gap, President Obama also reveals that he believes that hiring and placement discrimination is a major contributor to the gender wage gap, but in acknowledging that the Lily Ledbetter Act is only a small step toward eliminating the gender wage gap, he does see m to concede at least that some portion of
45 the gap is not caused by hiring and placement discrimination. Thus, similarly to Ledbetter Act do not answer. For instance, do gender wage gap extend to the portions of the gap that are not caused by hiring/placement discrimination? And if so, for this portion of the gender wage gap, where should the government draw the line in how far it will go to close it? The worth. Nothing t The fact that both the President and Congressional Republicans do not address these questions is understandable in tha t the Lilly Ledbetter Act itself is not intended to address every aspect of the gender wage gap, including to portion of the gap caused by comparable worth. But their failure to address them at the same time is very problematic from a policy perspective, because an informed policy response to the gender wage gap demands that they be answered. The fact that the Lilly Ledbetter Act was the first piece of legislation that President Obama signed indicates that, at least to the Democratic Party, the gender wag e gap is a top policy concern. These questions are addressed, however, by the CONSAD Report that was commissioned by the U.S. Department of Labor. Thus, the CONSAD Report holds special significance in that it speaks to many of the issues related to the g ender wage gap that were not addressed in the debate surrounding the Lilly Ledbetter Act, and it fact has implications about the
46 wisdom of the Lilly Ledbetter Act itself. To begin our discussion of the CONSAD Report, let us again review its findings: Although additional research in this area is clearly needed, this study leads to the unambiguous conclusion that the differences in the compensation of men and women are the result of a multitude of factors and that the raw wage gap should not be used as t he basis to justify corrective action. Indeed, there may be nothing to correct. The differences in raw wages may be almost entirely the result of the individual choices being made by both male and female workers (CONSAD, 2). As with Congressional Republica ns and President Obama, in drawing its conclusions the CONSAD report makes a normative judgment regarding the gender wage gap, even though it attempts to analyze the gender wage gap with a level of objectivity beyond what we see in the political discourse. Because of this normative judgment, the conclusion can only be unambiguous if every individual agrees with the normative judgment made by the authors of the report: that is, that if the gender wage gap is the The CONSAD Report finds that much of the gender wage gap results from women leaving the labor force to raise children, thus gaining less work experience than men and having lower levels of human capital, as human capital theory argues. Additionally, it finds that men tend to choose to go into careers with high wages, whereas women more often choose not to go into such careers, and it find s little evidence supporting the
47 notion that men and women choose different jobs because women are discriminated against by employers in hiring/placement. In my own review of the literature, I find that these claims generally are well supported by scholar ly research. Thus, the reason the CONSAD Report ultimately concludes that the wage gap does not justify corrective action is because if a woman chooses not to take time off of work to raise children, and chooses to go into a highly paying field, she will earn more or less the same amount of wage as a man who makes the same choices. And conversely, if a man chooses to take time off to raise children, or chooses to go into an occupation that pays a low wage, he would experience the same reduction in wages t hat a woman would choose. Essentially, the CONSAD Report is making the same argument that is put forward by Polachek (Polachek 26) in what he calls societal discrimination in which social forces are responsible for the economic disadvantage that women f ace. This argument implies that societal norms cause men and women to make different decisions regarding the division of labor in the home that favors the career aspirations of men, and additionally societal norms cause men and women to choose jobs that f it their gender identities. This is a clever argument against policies that would reduce the gender wage gap by targeting employers, which comparable worth policies would do, because it still implies that the gender wage gap is immoral and discriminatory, but essentially argues that do about it, because only a shift in cultural attitudes could really reduce the gender wage gap. I believe that the choices men and women make regarding their career paths do play an important role in perpetuating the segregation of jobs by sex. Similarly to
48 Polachek, I believe that these choices result from gender socialization that leads both women and men to view certain occupations as gender appropriate. However, whereas this acknowledgement leads the CONSAD Report and advocates of human capital theory implies that even if occupations are segregated because men and women choose different undervalued levels for such jobs are set lower in these jobs than in jobs that are traditionally dominated by men. This type of gender discrimination is distinct from other forms of discrimination, such as discrimination in hiring/placement. It is much harder to know that one is a victim of the discrimination in comparable worth, because of the surface men a nd women are being paid equally for doing the same job, but if the occupational makeup of that job happens to be mostly female, then it will pay less than a job that requires a comparable amount of skill and training that employs mostly males. Thus, while comparable worth does constitute a form of discrimination, and that discrimination is perpetuated by employers, the discrimination occurs at the occupational level rather than at the individual level. As such, it is a form of gender based discrimination that as of today remains legal in the United States. they are insinuating is that because women could enter more highly paying male jobs by tify government intervention because there is a way for women to increase their wages other than by addressing the gender based
49 discrimination against female jobs. But this does not address the fact that gender discrimination is still occurring if we be lieve in an egalitarian view of personal rights have to move into traditionally male occupations in order to increase their earnings rather, traditionally female occupations ought to have their wages increased to counteract the fact that these jobs have lower wages on the basis of gender. Additionally, if the type of occupational discrimination going on in comparable worth is occurring, then women moving into male jobs will only be l imitedly effective in reducing the gender wage gap, since women moving into male jobs could cause those formerly male occupations to experience a decrease in wages as their makeup increases in its percentage of female workers. Conclusion The notion that th e government should mandate that female jobs pay the same as male jobs when those jobs have comparable skill demands and working conditions certainly seems to clash with the neo classical view of how wages are determined. The neo classical view states tha t the wage of an occupation is determined by market forces that the market wage for a given occupation is determined by the level of demand for certain sets of skills, and the supply of individuals who possess those skills, and that employers themselves are price takers. Comparable worth seems as though it would usurp the role of the market in setting wages by giving the government that authority. For an economist who believes in the efficiency of market outcomes, embracing comparable worth policies may seem to come at the major cost of losing the efficiency of the market. However, I also believe that a responsible economist must recognize that
50 there are social forces that interfere with how markets function such that the supply and demand do not lead t o a truly socially optimal equilibrium. Discrimination is certainly one of these forces, and thus instead of viewing comparable worth policies that raise wages in female jobs as interfering with market wages, we ought to think of such policies as correcting interruptions in the market that lead certain kinds of work to be undervalued. The market wages for jobs that are primarily female are artificially low because the demand for workers in t hese jobs is artificially low due to the discriminatory tastes of employers who set market wages, and comparable worth policies are simply meant to capture the demand for workers in female typed jobs in the absence of discrimination. On the other hand, co mparable worth may also seem objectionable to some advocates of reducing the gender wage gap in that downplays the role that hiring and placement discrimination have in perpetuating the gender wage gap. This is not to say that hiring and placement discrim ination do not still occur, and that the government should not still make it a goal to reduce this type of discrimination. But it does imply that if we truly wish to reduce the gender wage gap, policy initiatives such as the Lilly Ledbetter Act that focus intentioned, are not going to be effective reducing the gender wage gap. Indeed, the fact that the gender wage gap reduced at such a rapid pace in the 1970s and 1980s but stalled in the early 1990s may indicate t hat legislation that focuses on hiring and placement discrimination reached its limit in its potential to reduce the gender wage by the early 1990s.
51 In this way, the theory of comparable worth seems as though it has the potential to be offensive to both ec onomists and sociologists in how it views the gender wage not that no one can support it, but that it shows how the division in the social sciences between economics and sociology is actually a fal se dichotomy. Their division is methodological as opposed to should be. We comparable worth policies bo th increase gender equality and make the economy healthier view justice and efficiency as a tradeoff, in the end we reduce both regardless of which one we choose to value more. E stablishing social justice increases efficiency and true economic efficiency therefore also increases social justice. This is the lesson within the lesson that comparable worth teaches us
52 Works Cited Albanesi, S., and C. Olivetti. "Home Production, Market Production and the Gender Wage Gap: Incentives and Expectations." Review of Economic Dynamics 12.1 (2009): 80 107. Blau, Francine D., and Lawrence M. Kahn. "Gender Differences in Pay." The Journal of Economic Perspectives 14.4 ( 2000): 75 99. JSTOR American Economic Association. The American Economic Review 84.2 (1994): 23 28. Budig, Michelle E., and Paula England. "The Wage Penalty for Mo therhood." American Sociological Review 66.2 (2001): 204 25. Burstein, Paul. Equal Employment Opportunity: Labor Market Discrimination and Public Policy. New York: Aldine De Gruyter, 1994. Cong. Rec. 22 Jan. 2009. http://www.gpo.gov/fdsys/pkg/CREC 2009 01 22/pdf/CREC 2009 01 22 pt1 PgS739 3.pdf. Web. CONSAD Research Corporation. United States Department of Labor. Employment Standards Administration. An Analysis of the Reasons for the Disparity in Wages Between Men and Women. Washington: 2009. Web. E ngland, Paula. Comparable Worth: Theories and Evidence New York: Aldine De Gruyter, 1992. Print. England, Paula. "Gender Inequality in Labor Markets: The Role of Motherhood and Segregation." Social Politics: International Studies in Gender, State & Socie ty 12.2 (2005): 264 88. England, Paula. "The Gender Revolution: Uneven and Stalled." Gender & Society 24.2 (2010): 149 66 L edbetter v. Goodyear. U.S. 550. Supreme Court of the United States. 2007 http://www.law.cornell.edu. Web. Leicht 255. Merrick, Molly K. "The Future of the Gender Wage Gap in the American Workforce." Honors College Capstone Experience/Thesis Projects Paper 281 (2010): n. pag. Web. Obama, Barack H. "Remarks of President Barack Obama on the Lilly Ledbetter Fair Pay Restoration Act Bill Signing." Office of the Press Secretary, 29 Jan. 2009. Web.
5 3 Perry, Jennifer, and David E. Gundersen. "American Women and the Gender Pay Gap: A Changing Demographic or the Same Old Song." Advancing Women in Leadership 31 (2011): 153 59. Web Polachek, Solomon W. "How the Human Capital Model Explains Why the Gender W age Gap Narrowed." Discussion Paper No. 1102 (2004): 1 43.