Diploma in Business Management

Developing compensation programs that have as a major goal the creation of equity in pay has been a focus in this book. This final chapter deals with equity as it applies to the problem of discrimination in pay, particularly toward women.

Attaining equity in pay is a difficult goal to achieve, for a number of reasons. The first is that equity is cognitive: it depends on the perceptual field of the person for its definition. It may vary between the two parties in the exchange or with observers to the exchange. Second, equity in pay may conflict with other goals of compensation. In particular, equity is often seen as conflicting with the competitiveness of the organization in the marketplace. Third, the legal environment, while encouraging equity in some ways, makes it a burden and defines the process to be followed in cumbersome ways. Fourth, the organization may find itself as a leader of social change rather than a follower. This makes the perceptual problem just cited particularly difficult to deal with, for different groups of employees have very different ideas about what is equitable. When one group sees that equity is being gained at its expense, the organization is placed in the middle.

In a way, consideration of discrimination in pay is an issue whose time has come. For the past 40 years, organizations have been made more aware, through legislation and pressures from civil rights groups, of the way in which they discriminate against certain groups in their Human Resource practices. But the major thrust of this pressure has been on employment practices, called access discrimination. Although no one would claim that access discrimination does not exist today, there are clear guidelines as to how organizations are to act. A second kind of discrimination is valuation discrimination. This type of discrimination occurs when the employee�s contributions to the organization are not properly rewarded because they are a member of a protected group. Discrimination of this sort has not received the consideration that access discrimination has, so the level of sophistication in dealing with it has not been reached. It is not as clear whether there is a problem or, if there is, what action organizations should take to solve it. This chapter examines how to identify the problem of valuation discrimination in organizations and describes ways in which organizations can deal with it.

Discrimination in pay is inevitable and indeed desirable in organizations. All organizations pay differently for different jobs and in consequence pay people different amounts of money. Even if an organization were to pay all employees the same, some employees would unquestionably claim that this kind of equality is discrimination. So the problem is not whether to discriminate in pay but rather how organizations go about discriminating.

Compensation, as a reward in organizations, is intended to create behaviors that lead to the accomplishment of organizational goals. When differentials in payment are based upon this principle, then the organization may be seen as discriminating in the best sense of the word. In this book it has been suggested that the basis of pay differentials should be the nature of the job, the performance of the job, and the personal attributes that enable the person to do the job.

In the Equal Pay Act, four factors are defined as appropriate differences: skill, effort, responsibility, and working conditions. These four are similar to the three just cited, although working conditions adds the idea that the context within which work is done may be a reasonable factor. Most employees perceive establishing pay differentials on the basis of these job-related factors as reasonable and fair. The way in which they are administered, however, may lead to feelings of inequity, as will be discussed later in this chapter.

Discrimination can be improper in two senses. The first is when the differentials in pay are not related to factors that lead to behaviors on the employee's part that are in turn related to the accomplishment of organizational goals. In this instance, it is the organization that is not receiving its money's worth when it pays its employees. Second, these irrelevant criteria may harm the employee. Paying employees different pay at random is senseless, and the employee is likely to feel that it is just luck as to what he or she is paid. But paying one employee less than another on the basis of sex is a defeating experience for the lesser paid person, since sex is not a characteristic that one can change and is not perceived by the person as relevant to performing the job.

Legally, using certain factors to differentiate pay is illegal. The Civil Rights Act says that it is unlawful to discriminate against any individual with respect to compensation, terms, or conditions of employment, because of such individual's race, color, religion, sex, or national origin.

These improper factors are quite clear, but what is not clear is when these are the factors that are being used. If women are paid less because they are women, then the wrong criterion for pay differentials is being applied; but if women are paid less because of the job they hold, then the question of whether this is improper discrimination is not as clear.

Improper discrimination in pay by organizations appears to occur mostly on the basis of sex. The feminist movement originally focused attention on this by popularizing the idea of the 59 cent dollar. By this they were pointing out that women earned 59 cents for every dollar of income that men earned. Today that figure has risen to the 75 cents dollar. This is better but shows that there remains more to do. This difference clearly represents a significant earnings gap between male and female workers.

The question, however, is why there is such an earnings gap. Is it that organizations consciously discriminate and pay women less than men, or is this an unplanned result of establishing pay differentials based on rational criteria? This is a complex question, and this section examines it from three perspectives: economic, legal, and psychological, in order to see how we arrived at this state and what is likely to happen.

Economists have been very interested in studying the earnings gap between men and women. Their approach has mainly been to examine the problem from a macro perspective, attempting to see if there are differences in men and women workers overall that account for the lower average wages paid to women. Their studies indicate that a majority of the differential can be accounted for by a number of factors, discriminatory and nondiscriminatory, which we will turn to next. When these factors are taken into account the gap between men's and women's wages narrows to around 12 cents.

Participation rate. The first and most dramatic factor affecting women's wages today is the increased participation rate of women in the work force. There has been a steady increase in this rate since after World War II, when around 45% of women worked, to today when over 60% work. This change may well be the most dramatic change in the labor force in this century, but it has occurred so gradually that it has been called the Subtle Revolution. The expected effect of an increase in the supply of workers of this magnitude is to depress wages for that group of workers, and this has undoubtedly happened. In addition, with the recent changes in welfare there has been another large increase of women into the workforce, this time with fewer skills and experience.

The reasons for this increase in participation are many and involve a drastic change in American society. Lower birth rates and improved technology in the home have made it easier for women to work. Increased education has made it more desirable for women to work. But the need for two paychecks in many families and the increase in the number of one-parent families, usually headed by a woman, have made it necessary for women to work. The single adult family accounts for a rising proportion of America's families, particularly poor families. This trend has been called the feminization of poverty.

Job segregation. The increase in the participation rate might not by itself have reduced women's wage rates. If there were a single labor market, then an increase in the participation rate would simply mean an overall increase in the supply of labor and somewhat lower wage rates for all workers. But there are in actuality many labor markets, and women tend to be segregated into only a few of them. Figures show that whereas 50 percent of male workers have jobs in 63 occupations, 50 percent of women workers are concentrated into 17 occupations. In order to eliminate this segregation, it is estimated, almost three-quarters of all working women would have to change their occupations. Further, the occupations occupied by women are lower-paid than those dominated by men.

Job segregation of women is not limited to occupational areas. Women tend more than men to enter industries that pay lower wages overall. Even when women are in male-dominated occupational areas, the specific jobs they hold are at the bottom of the wage ladder in that occupation. This may be partially a case of employment discrimination and partly because the increased participation of women means that they are young and concentrated in the beginning steps of an occupational ladder. For those occupational areas in which women predominate there is one further problem: there is very little vertical movement within the occupation, and the range of salaries is therefore narrower.

Job segregation between men and women is a deep seated sociological phenomenon. There are appropriate roles for men and women in all societies. In the United States there is a blurring of these roles with increased participation of women in the work force, changing education patterns of women, and the effect of civil rights legislation The changing nature of the economy, from manufacturing to service, encourages this trend. Where men and women are working on the same job, there is a dramatic narrowing of wage differentials, providing little evidence, either at the economy level or within organizations, of pay discrimination.

Education. Today, more women than men attend college. Although females are still under-represented in the fields of computer science, physics, and engineering; females do make up more than 40% of medical and law students.

Experience. Experience levels of women are affected by the tendency to drop out of the labor market for childbearing. At a minimum this means women have less experience than men the same age. In addition, dropping out may mean a lessening of skills in occupations that are changing rapidly. Further, employers are less likely to engage in training personnel whom they perceive as not being permanent employees, so even if women have the same level of experience they may have less training.

Labor Market. The above factors, particularly the increased participation rate and job segregation, are likely to lead to a situation in which women's jobs are paid less in the marketplace. Since organizations need to be efficient, they should not pay any more than they have to for any resource, labor included. Viewed in this way, paying women less than men is simply following what the market dictates. It is claimed that administrative interference with wage setting would lead to inefficiency, and would not work. For instance, in Australia there appeared to have been a rise in female unemployment caused by raising women's wages artificially.

As this book has pointed out, however, the labor market is not a perfect vehicle for determining wages. The going rate is not a point but a range of wages. Deciding what is the going rate is in itself an administrative decision. But most important, establishing wage structures is a process of integrating both internal and external wage pressures. Often it is the internal considerations that dominate in the wage-setting process. So the market is not a totally dominant force but rather one of the considerations the compensation specialist needs to keep in mind.

The legal perspective on discrimination in pay is based upon two doctrines, equal pay and comparable worth. The first is a product of the Equal Pay Act of 1963 and is well established in law. The second is based upon the Civil Rights Act of 1964 but is not well established in law at this time. This section will discuss each of these doctrines.

Equal in this law means the same as or, more accurately, substantially similar. That is, the work activities performed by male and female workers must be the same or almost the same. Where there are some activities carried out by males occasionally but not regularly, this is not a basis for a differential in pay. Further, it is the actual activities and not the formal stated ones that are considered. This means that the job descriptions of the organization need to be kept current and reflect the activities of the employees in those jobs. The appropriate technique for evaluating equal pay is job analysis.

The determination of whether two jobs are substantially equal is based upon the four factors mentioned above in the Equal Pay Act. Guidelines for interpreting the law have evolved from court cases. The skill, effort, and/or responsibility involved in a job must be substantially greater than in another job for the two jobs to not be equal. The tasks involving this greater degree of the above must constitute a significant amount of the time of all employees in that job classification. The differential skill, effort, and responsibility should have a value commensurate with the pay differential that is in question. A further indication is different recruiting practices or training for the two jobs.

All in all, organizations must pay equal wages or at least have equal pay ranges for jobs that perform substantially the same work. It is the actual work that is done that controls this definition. This puts a weight on performing good job analysis in the organization, keeping job descriptions up to date, and having the compensation staff know what is happening in the organization.

Comparable worth. The concept of comparable worth takes the idea of equal pay for equal work one step further, comparing not only jobs that are the same but jobs that have the same value to the organization � that is, jobs that are comparable. This idea is not included in the Equal Pay Act, but proponents claim that it is inherent in the Civil Rights Act of 1964. As noted, it is unlawful to discriminate on the basis of sex and race as well as other factors in any terms of employment, including compensation. This is a much broader prohibition than that in the Equal Pay Act. At this time the courts have said that the Civil Rights Act may be invoked in compensation cases, but exactly how and to what degree are not clear.

Comparable worth has caused much more argument than equal pay. Proponents point to the 75 cent dollar and claim that a major part of this discrepancy exists because women's jobs are undervalued by society in relation to men's jobs. Proponents feel that comparable worth will be served by comparing men's and women's jobs using job evaluation, and that if this is done it will show that women's jobs are indeed undervalued by society. In fact, this has been done in Washington State, Los Angeles, and San Jose, and the results show there to be an undervaluing of women's jobs.

Thus, proponents of comparable worth point out that not only are women segregated into a few women's jobs but that because women occupy these jobs they are downgraded. Opponents answer that women can move into other jobs if they so wish. Proponents, in turn, point out that the training of women is provided in these women's jobs, that these jobs are intrinsically satisfying to women, and that women should not have to enter job arenas that are not as satisfying in order to gain pay equity.

Opponents of comparable worth have called it "the looniest idea since Looney Tunes," and this was from the chairman of the U.S. Civil Rights Commission.They contend, accurately, that the idea was rejected in the discussions that led to the passage of the Equal Pay Act and that Congress therefore did not wish to go this far in reducing inequality. The main problem that opponents see in comparable worth is that it ignores the labor market. They perceive organizations as merely paying market rates for jobs, and to do otherwise is to destroy the market mechanism. Further, the idea seems to them to be impractical. They see comparable worth as involving four elements that are objectionable and hard to achieve: (1) comparing dissimilar jobs, (2) comparing these dissimilar jobs on their so-called intrinsic value, (3) developing an unbiased job evaluation plan, and (4) requiring third-party intervention to make these determinations. In short, opponents see comparable worth as impossible in the labor market and impractical in administration.

As in most emotionally charged issues, neither side is totally accurate. The proponents see the undervaluing of women's jobs as the major contributor to the 75 cent dollar. Given the analysis in our discussion of the economic perspective, this is not very likely. Part of the problem is undervaluation, but a fully installed comparable worth program would not create equality of wages between men and women. Also, reliance on job evaluation ignores the fact that job evaluation does not create relative value of jobs in dollar terms, nor is there a conversion system other than using market rates.

On the other hand, opponents underestimate how much job evaluation can do. The argument that dissimilar jobs cannot be compared is contrary to the whole process of job evaluation in organizations. Likewise, the idea that organizations passively pay the market rate is inaccurate. As seen in this book, developing a wage structure is a process of integrating internal organizational value with the market value, and which is dominant depends on the jobs and the organization. Further, the measurement of the market rate is also a judgment that can influence the stated rate.

The legal status of comparable worth is not as clear as that of equal pay. The Civil Rights Act is a much broader act than the Equal Pay Act and has required a great deal of court interpretation in all areas. In the area of comparable worth, there are likely to be more decisions in the future to define the limits and establish the process. At this time the following points seem to be established:

1.	The Civil Rights Act may be used in a case of disparate compensation between men's and women's jobs.
2.	The disparate impact standard that is commonly used in civil rights cases can be a starting point, but it is necessary that intent also be shown. Where widespread sex discrimination has been shown in the organization or where the discrimination is blatant, comparisons of dissimilar male and female jobs were allowed, but discrimination caused solely by market forces is not a cause for action.
3.	Organizations must take action when they discover that there is discrimination in pay. Ignoring the results of studies showing that women's jobs are undervalued is a basis for legal action.

In summary, there is likely to be continuing pressure to examine discrimination in pay within organizations from a number of sources � feminist groups, unions, legislatures, and the courts. Organizations are going to have to defend the differentials in their organizations as being equitable.

Our discussion of comparable worth would indicate that there are considerably different perceptions about whether there is a problem of pay discrimination, about its nature, and about actions that should take place. These perceptions are likely to influence how people, in this case women, react to their compensation. Analysis of these reactions can be done by using the two models of motivation discussed in this book, the membership model and the performance model.

Membership model. The membership model of motivation is based upon equity theory. This theory states that membership is a perceptual balance between the contributions required of the person and the rewards received by that person. If these are not in balance, the person will take action to alleviate the resultant cognitive dissonance. In the case of pay discrimination of women, it would appear that there has been a change in their perceptions that has created or increased their dissonance and that this dissonance will not go away without action taken by employing organizations.

Individuals feel equitably paid when they perceive that their rewards from work equal or exceed their contributions. In the case of women, there was in the past a tendency for them to undervalue their contributions and overstate their rewards. Women were likely to agree with the cultural norms and feel that their contributions to organizations were not as great as those made by men in terms of both the types of job contributions and the types of personal contributions, such as skill and training. In addition, they tended to increase their non-pay rewards by emphasizing the good working conditions of the office vis-à-vis the factory floor and the social aspects of the job.

This set of perceptions is changing. Women who are primary breadwinners are less interested in the non-pay aspects of the job and more interested in the pay they receive. The traditional position of not making more money than their husbands is disappearing: there often is not a husband and where there is this is less likely to be a consideration. As women's consciousness has been raised the value they place on their contributions to organizations is rising. They are now more likely to perceive that the skills they bring to the job are as valuable as male skills. However, organizations have not changed their perceptions as fast, leading to a situation in which women find that they have changed but their work situation has not, thereby creating feelings of inequity.

Today more women have the same background and skills as men, and those who have more traditional skills feel that those skills are as valuable as men's skills. The result is that women are more likely to try raising their pay in order to reduce their dissonance.

This pressure is not likely to go away, since women today are less likely to accept the traditional norms about women's work. In addition, organizations can expect women to now make men's jobs the comparison by which they judge whether they are being fairly rewarded. This puts organizations in an uncomfortable position. To the extent that men and the organization maintain the same perceptions that they have in the past, there is going to be tension between women and their employing organizations. This is the type of situation that has led to unionization of groups in the past, and unions are today attempting to take advantage of this disparity in perception.

Performance-motivation model. Changes in perception are likely to affect the motivation to perform as well as create feelings of inequity. As we have seen, the performance-motivation model has three parts: the value of the reward, the performance-reward connection, and the performance? effort connection.

The value of the reward, pay in this case, appears to be going up for women. This should make it easier for organizations to get women to focus on working harder on the job, provided the other two parts of the model are fulfilled. The performance-reward connection is difficult in many women's jobs since women are more likely to provide a service rather than a product. In the past, there has been a tendency to shy away from making direct performance-reward connections in women's jobs. There appeared to be a feeling that they were inappropriate for that group of employees. As women move into men's jobs this distinction lessens, but there are many women's jobs for which establishing a performance-reward connection is just beginning. Nurses and teachers are two such groups.

The performance-effort connection has also been tenuous in women's jobs. This is due partly to their service type of work; again nurses and teachers are good examples. In neither of these areas is the person sure what impact she has had on the outcome. In fact, women in both professions find this a frustrating part of their job. There is also the problem that in women's jobs very often the performance-effort connection is purposely reduced by the organization; the woman is thereby not allowed the discretion necessary to make the connection.

This analysis of the psychological perspective leads to the conclusion that there have been some clear changes in the perception of women about their jobs and their compensation. Organizations are in a position to move toward employing, the performance-motivation model more clearly in women's jobs with an expectation that women will respond positively to these changes.

Until now we have examined the problem of pay discrimination from a societal viewpoint. This section looks at the issue of pay discrimination within the organization and makes suggestions as to how to find out if there is improper discrimination, either overt or unconscious, and how to improve compensation decisions so as to eliminate or at least reduce improper discrimination. We examine the three major compensation decisions-wage level, wage structure, and wage system.

Wage level decisions

The wage level decision has to do with who gets what, so it is very apropos when discussing discrimination. Wage level was defined in this book as the "average wage paid to workers at some level of analysis." In this case the level of analysis is the male and female employees in the organization. In most organizations this analysis will produce results like that for the society as a whole � that women are not paid as much as men in the organization. Whatever the size of the differential, its reasons need to be examined. The first level of analysis is the difference in male-female wage levels in parts of the organization and by job clusters. This analysis should reduce the differential. At the end this analysis should have been applied to each job category in the organization, and within these categories the only differences should be based on those criteria expressed in the Equal Pay Act � seniority, performance, and differences in quality or quantity of production.

The economic perspective discussed is a guide to why there may be a difference between male and female wages in the organization. Job segregation is a problem that has been brought to the attention of organizations by Glass Ceiling Analysis. The purpose of this type of analysis is to examine where women are placed in organizations, how to get them into male-dominated job categories, and how to move them up in the organization to high levels of responsibility. This initiative is mainly one of access discrimination but has aspects of valuation discrimination, requiring organizations to examine male and female wages by organizational levels.

Wage level policy for job groups. Organizations usually maintain different wage level strategies for different job groupings. These strategies are associated with the importance of the jobs to the organization. In an engineering organization, the wage level strategy is most likely to pay engineers somewhat above market while paying market or below to other jobs in the organization. They do this to ensure that the human resources most valuable to the organization, engineers in this case, are readily available and are the best that can be found. This differentiation of strategy among job groupings can exacerbate the discrimination problem where the jobs paid above average are male jobs and the jobs paid below average are female jobs. A common type of differentiating strategy is illustrated in figure 26-1. In this situation, the organization has determined that it wishes to pay above average for the top jobs in the organization and below average for those at the bottom of the organization.

If women predominantly occupy the lower-level jobs in this organization, then the differential between men and women will be enlarged by such a strategy.

Wage surveys and the labor market. The largest disagreement between the proponents and opponents of comparable worth lies with the function and use of the labor market as the determinant of wages. Opponents claim that wages should be whatever is the market rate. Proponents claim that the market rate for women's jobs is discriminatory and that therefore market rates should not be used. Both these positions are all-or-nothing positions. In fact, market rates are used for setting the value of jobs, but not exclusively; and jobs differ in the degree to which it is useful to try to determine market rates. So market rates are one factor in determining the wage rate for a job but not the only one. In some cases, this use of the labor market can increase the differential between men's and women's jobs. In performing a comparable worth study, the rates for male dominated jobs are used to establish the pay line. Then wages for women�s jobs are examined in light of the men�s wage line.

The wage structure is a combination of the market rates as determined by the wage survey discussed above and the job structure of the organization. In turn the job structure is determined by job analysis and job evaluation. These processes can both help and hinder efforts to reduce wage rate differences between men and women.

Job analysis. Job analysis is the technique that is used in determining equal pay. Figuring out if two jobs are substantially the same involves finding out if they have the same tasks. This is a descriptive process, not an evaluative one. Since job analysis is supposed to be a descriptive process there should be little chance for discrimination to take place. But since humans collect the information there is always the possibility that the sex of the job incumbent and/or the job analyst can influence the data collection. This possibility has been investigated, with the conclusion that there does not appear to be much discrimination here. The findings of these studies indicate that the amount of information is a major variable in whether the sex of the person being analyzed enters into decision making. From this it appears that it�s necessary to make sure that the job analyst collects sufficient information.

In a different approach, one of the authors examined whether men and women described their jobs differently. Incumbents of the same job title were asked to describe their activities, decisions, and interactions. The responses were content-analyzed. Major differences were found in the words men and women used to describe their activities, and these differences were found to be detrimental to women in that the words women used were evaluated lower. It may be important, then, to look at how job descriptions have been developed for women�s jobs to see if they are ready for job evaluation.

Job evaluation. The question in the relationship of discrimination in pay and job evaluation is whether job evaluation is the solution or part of the problem. It appears to be a bit of both. A majority of organizations use some sort of job evaluation to establish a job structure. They most often use job evaluation because the market rates are unavailable or hard to obtain or because they wish to pay other than the going rate for some of their jobs. But job evaluation leads only to a job structure and not to a wage structure. There is no inherent worth developed through job evaluation. What is established is a hierarchy of jobs within the organization which is then referred to the market values in order to establish pay rates for the job. So job evaluation, by itself, cannot establish the monetary value of the job to the organization or the economy.

However, job evaluation can develop the relative positioning that the organization establishes for jobs and can compare this positioning with that of the labor market. Where the two disagree the organization must determine whether it wishes to pay market rates or not. This decision does not always go in favor of the market rate, although it must take into account the effects of not paying the market rate. Organizations have different needs and values than the general economy does at a particular time. Thus, they may pay higher or lower than the market in order to most effectively pursue their own goals. Reducing the differential between male and female wages in the organization can reasonably be one of those goals.

In summary, job evaluation is a tool that can compare apples and oranges, which the opponents of comparable worth say cannot be done. Organizations can and do have job evaluation plans that rate all jobs within the organization. On the other hand, job evaluation cannot replace the use of market rates in setting wages, as is proposed by the proponents of comparable worth, since the results of job evaluation are not stated explicitly in monetary terms, nor can any convenient ratio be assumed. Job evaluation can be used to rate and/or rank all organization jobs and to compare that ranking with market rates to establish the wage structure. As stated above, the proponents of comparable worth then want the wage structure to be developed by using the labor market for male-dominated jobs.

Although job evaluation can be used as a partial solution to the pay discrimination problem, it is also a part of that problem. Job evaluation is a judgmental process and is subject to the human errors that can exist in any such process. One of the first steps in job evaluation is determining the compensable factors to be used. Do some compensable factors discriminate against women, or are compensable factors that are important for women's jobs ignored in job evaluation plans? One study has shown that for some compensable factors the evaluation of male and female jobs did affect the ratings. But this effect was not strong. However, this is a possibility that has not been thoroughly explored. Some important aspects of women's jobs may tend to be ignored in job evaluation. Interactions with others may be a case in point. Many women's jobs require considerable contact with others. This factor is ignored in most job evaluation plans and may be biased toward the kinds of interactions typical of men's jobs.

A second question relates to whether the type of job evaluation used makes a difference. Early studies comparing the results of job evaluation methods showed little difference between results obtained from different methods. Two recent studies, one of which focused on the impact of different job evaluation plans on pay discrimination, show a considerable difference in the results of job evaluation plans. Most of these differences were to the detriment of the evaluation of women's jobs.

The third area where discrimination is likely to occur is in the actual evaluation of jobs. Most job evaluators, after they become experienced, find themselves using the formal plan to validate their initial impression of the job. To the extent that this impression is a result of knowledge of the job and the evaluation instrument, the results are probably accurate. But if the information is inadequate, as we have discussed, then the results are likely to be flawed. One study examined the influence of current wage rates on the evaluation process and found a significant effect. Thus, the fact that women's jobs are paid less results in their being evaluated lower than men's jobs. All these criticisms indicate that job evaluation is not a perfect tool for determining whether women's jobs are underpaid vis-à-vis men's jobs. But since organizations have been using the technique with success for so long, it is reasonable to assume that job evaluation results can be one way of examining discrimination in pay.

Wage systems decisions have to do with determining the wage rate of the employee within the guidelines of the wage structure. It is at this point that the personal contributions of the individual enter into the wage determination. The major discussion of pay discrimination has focused on jobs, but there is another series of studies that have examined the contributions of women in the workplace as compared with those of men. Economists claim that one of the reasons that women are paid less is that their contributions, defined as human capital, are less than men's. But are women's contributions valued as highly as men's? There are studies indicating that women and other minority groups earn less than white males of equivalent age, experience, and education. This problem can be studied within the organization by developing a regression equation of the salaries of males based upon their age, experience, education, and any other relevant characteristics and using this equation to predict female salaries. For more information on this technique see DLC Course 49: Regression Analysis Used in Compensation Administration. This prediction can then be compared with women's actual salaries. The results might look like figure 26-2.

Source: M. W. Gray and E. L. Scott, A 'Statistical' Remedy for Statistically Identified Discrimination, in Academe, May 1980, p. 177.

Given this discussion and the process of individual pay determination, the examination of discrimination in wage system decisions will focus on three areas: where the employee started in the pay range, how long the employee has been in the pay range, and the performance of the employee.

Starting wages. This may be an overlooked area in which pay discrimination occurs. The wage rate that a person accepts on his or her first job, and especially the first job with a particular organization, influences the wage rate of the employee from then on, since increases in pay are most often a percentage increase over current wages. In the past, there was a tendency to offer women not only lower-level jobs but lower wages on the same job. Although this practice is illegal under the Equal Pay Act, it is very hard to detect, since job offers are clearly made within a particular pay range. The organization would need to keep track of the average pay offer for each job category for men and women to control this tendency. Women exacerbate this by being willing to accept lower pay than men for equivalent jobs. This may be particularly true for women reentering the work force after being out to raise children, for they may feel unsure of their talents at the time. Once they are back working and feel confident, they are still saddled with the decrement created by the salary at which they began.

One study in this area showed that women were offered lower wages on average than men for a particular job. To compound the error, they were more likely to be assigned more routine tasks and then given lower second-year increases than men. A court case that illustrates the problem of women getting behind men in wages based upon prior pay history is Kouba vs.Allstate. In the 1982 Allstate case women who were selected as sales agents from within the company had starting salaries lower than their male counterparts since they had come from lower-paid jobs. The trial court agreed with the women that the jobs they had come from were probably subject to discrimination and that the company should pay them on a par with the male sales agents. The appellate court, however, agreed with Allstate that its actions had a business reason and not a sex-biased reason. Although Allstate won this case, it brings to light the problem of the variation in pay history between men and women.

Time in grade. Two factors can hinder women in regard to time in grade. The first is the turnover rate. Many women�s jobs have a high turnover rate, so few women get to the top of the range. The affect on the average wages paid to men and women in the organization is to lower women's average wages. The second factor is that women's jobs tend to be at the bottom of the wage structure, which has pay ranges that are narrower than those at the top of the wage structure. So even if they did stay in the job they would not be able to move up as much as men.

Performance. Discrimination can occur in a pay-for-performance plan if women's performance is viewed differently than men's performance. Studies on the effect of sex on performance ratings have varied widely in their results. Some show no effect, some a positive effect for men, and some a positive effect for women. The one thing that seems to account for these variations in results is the amount of information. That is, people rely on stereotypes when they do not have clear and objective information on which to make a judgment.

There is evidence that women rate themselves lower than men and that women rate others higher than men do. Thus, women would be more likely to accept lower ratings than men, and supervisors would be more likely, therefore, to feel free to give lower ratings. Also since there are fewer women supervisors, their tendency to rate higher would not have a strong effect within the organization. However, one recent study did not support either of these contentions and showed both men and women rating themselves more highly than did their supervisors.

Last, there may be some concern with the direction of performance rating and pay determination. It is probable that current pay affects the determination of performance. In order to be consistent, supervisors in a pay-for-performance system are likely to rate employees high or low, based upon where they are in the pay range presently. This smooths out the percentage increases that they give out to employees.

The responsibility for dealing with pay discrimination will fall on organizations in the end. Legislation and court decisions may provide guidelines, but the work will get done in individual organizations. The areas that the organization can focus on to reduce discrimination are compensation-plan design, administration, and control.

Comparable worth is seen basically as an internal equity problem. This suggests that job evaluation will always be important in determining whether discrimination is taking place in the organization. Most of the suggestions in the area of job evaluation are for the organization to have a single job evaluation plan for all jobs. This is not an uncommon practice, but it should be looked at seriously by those organizations using two or more plans. An alternative is to use key jobs to compare the plans. By evaluating these jobs in each plan, you can determine if the results are the same.

The job evaluation plan itself needs to be designed so as to minimize discrimination. The choice of compensable factors can help: factors that are clearly important to both men's and women's jobs would be preferable. Also suggested are plans that rely on more sophisticated procedures, such as factor comparison and point systems that are likely to give more accurate results and reduce the opportunity for intentional discrimination.

Although external considerations appear secondary in comparable worth, it is necessary for the organization to minimize the discriminatory effects of the labor market. This can best be done in developing the pay-policy line by using key jobs that are based upon male jobs or a combination of both male and female jobs and using this single pay line in determining the value of all jobs.

Pay systems would be aided greatly by staffing and career changes in organization. Career paths for women's jobs need to be opened up so more movement can be made, but most important is to help reduce the job segregation of women by placing them in male-dominated jobs. This is seen by most people as the best way to reduce pay discrepancies. Performance appraisal must also be designed to focus on behaviors that are important in both male and female jobs.

Administration of compensation plans deal with a multitude of decisions made by many people in the organization. As with access discrimination, it is necessary to develop policies and procedures to guide this decision making. As indicated, the amount of information provided to decision makers' is a major variable in whether they will use sex as a consideration in compensation decisions. Thus, it is imperative for the compensation staff to ensure that all the proper information is available and used in making pay decisions. It is particularly important to have sufficient information available for performance appraisal decisions, and for setting pay rates for new hires.

A quote from the report of the �Glass Ceiling Analysis� is apropos to initiate this section: �When an activity is measured, monitored and reported, there is a much greater incentive by everyone involved to show positive results.

This is the lesson of control systems, people pay attention to what is measured. For many reasons, organizations need a compensation control system, not the least to control the largest cost of most organizations. A great deal can be gained by developing a compensation audit to analyze the possibility of bias and discrimination in the compensation system. The OFCCP has developed a process for such an analysis for government contractors, but any organization can benefit from their analysis.

Audit. In order to perform a compensation audit two sets of information are required:

Data analysis. Analysis of this data may be done at a number of levels; from the simplest to the more complex they are:

1.	Developing tables according to the categories above and then sorting them by certain ways, such as:
		a.	sorting by salary to find the highest paid woman and/or other minority employees in the company
		b.	sorting by hire date to find out if certain groups have longer tenure
		c.	sorting by departments to see if women and minorities are concentrated in particular areas
		d.	two other types of non-statistical analyses are:
				i.	cohort analysis: in this case protected (female) employees are matched to non-protected (male) employees in terms of the categories developed above as important in pay decisions. Wage rates of the cohorts are then compared.
				ii.	Outlier Analysis: this analysis focuses on pay grades and examines the very high and very low employees in each group to determine if they fall predominately into either the protected or non-protected group.
2.	The first level of statistical analysis is Median Analysis. This can be done by arraying the salaries of all employees in the chosen category(s) and arranging them in ascending or descending order. The middle salary is the median. The median is particularly useful as it (1) can be done without any mathematical calculation and most important (2) is not influenced by one or more extreme salaries. Medians that are substantially different for protected and non-protected employees within the category need to be examined to see what explains the difference. Not all differences are discrimination; some may result from differences in something like performance levels.
3.	The second level of statistical analysis is calculating the average or arithmetic mean. See Chapter 5 of the text . The mean provides that average of a series of wages, in this case. The means for the protected and non-protected employees can then be compared. The advantage of using the mean is that the two means can be compared using statistical processes, such as T-tests.
4.	Having two means that are statistically different indicates that there may be discrimination. What is needed is a technique that can explain the nature of the differential. This is the role of regression analysis. Regression analysis looks at all or selected factors that your company uses to set wages and regresses these against the array of salaries for a category. If these factors explain the variance then there is presumed to be no discrimination. However, if these factors do not explain the variation, then it is presumed that discrimination is present.