Ellen Isaacs: Gender discrimination in the workplace

This is a short article I wrote as a sidebar to a much more thorough (and thoughtful) article by Kathy Hemenway called Human Nature and the Glass Ceiling in Industry that was published in Communications of the ACM in 1995, Vol. 38, Number 1, pp. 55-62. A PDF version of Kathy's paper is available, and it includes my sidebar on page 58.

This survey was prepared for an industry task force report on women in software engineering. It was intended to supplement that report by providing context on gender discrimination in the workplace with a focus on software engineers.

Introduction

Data collected from a wide array of sources reveal the following pattern with regard to gender discrimination in the U.S. workplace. In general, the proportion of women employed as computer scientists appears to reflect the proportion of women graduating with degrees in that area. However, when women are hired, they tend to start at lower positions and/or earn lower starting salaries than men. Over time, the gap between men's and women's salaries and promotion rates grows at an increasing rate. The salary gap is found even in studies that equate years of experience, level of education and industry.

Hiring

There is some evidence that the proportion of women computer and mathematical scientists hired into industry jobs reflects the proportion of women graduating with degrees in those areas. Data from different sources indicates that women made up about 30% to 35% of all computer and mathematical scientists between 1988 and 1990 (Frenkel, 1990). This figure is somewhat higher than the proportion of women graduating with CS degrees over the past 10-20 years. According to the U.S. Department of Education, women earned between 30% and 37% of the bachelors degrees in CS during the 1980s, up from 14% to 28% in the 1970s (DOE surveys, 1990-1991). As for higher degrees, women earned between 21% and 30% of the CS masters degrees in the U.S. between 1980 and 1989 (Frenkel, 1990), and they earned between 9 and 14 percent of the CS PhDs in the U.S. and Canada between 1978 and 1990 (Gries and Marsh, 1992).

Promotion

In almost every industry, women occupy a very small proportion of the higher-level positions. For example, a 1988 study found that only three CEOs among the Fortune 1000 were women, and only 1.7% of the COOs, CFOs and executive VPs were women (White, 1992). In a 1993 study of Stanford MBAs, graduates from the class of 1982 were tracked over time. It was found that 71% of the men are currently in the top four rungs of management, whereas only 34% of women had reached those positions (Smith and Mitchell, 1993). A study of the 10 largest makers of weapons found that women made up 5.3% of the senior management positions (Sims, 1993). Business Week did a report in 1987 in which they tracked 100 women executives who were on the fast track from as far back as 1976. They found that none of those 100 women had made it to the top position in a public corporation unless they started the business or inherited the position (White, 1992).

Data from the computer industry in particular were not available, but the same pattern appears in the academic world. Women make up 10% of both assistant and associate CS professors but only 4% of the full professors, a rank that generally takes about 10 years to achieve (Gries and Marsh, 1992). (Recall that women have been earning between 9 and 14 percent of computer science PhDs in since 1978.)

As shown in the previous section, the problem is not that larger proportions of trained women are not available. Women are not represented at the highest ranks of companies and academia because, for some reason, their rate of progression is halted somewhere along the way to the top.

Salaries

The salary picture for women is even more inequitable than that for promotion. Women consistently make less money than men in almost every industry, even when they first start their jobs (Schwartz, 1988, Mahar, 1993). An American Demographics study found that women working full time with two or fewer years of experience earn 72% of men with the same experience (Schwartz, 1988). In the computer and mathematical sciences, women's wages as a percentage of men's has fluctuated between 74% and 86% between 1983 and 1992, although on the whole it has grown from 75% to 85% (U.S. Department of Labor statistics).

As women get older, they make less as a proportion of men's salaries. Although the gap has narrowed somewhat in the past 14 years, this trend is due to a drop in men's inflation-adjusted salaries, not a rise in women's (Pennar, 1991, Rigdon, 1993). And the gap has not been steadily decreasing. In 1955, women earned more of a percentage of men's salaries than they did in 1987, 63.9 cents vs. 63.7 cents (Mahar, 1993).

Part of the reason for the wage gap is that women don't get promoted as quickly as men. However, even when equating for rank, a gap appears. A 1993 ComputerWorld survey of IS managers salaries showed that the wage gap widened as the management level increased (Dwyer, 1993). For example, among " programming managers," women made 98% of men's salaries, but among IS directors or managers, women made 82% of men's salaries.

Other common explanations for the increasing wage gap are that women choose professions that pay less and that they have less experience than men of the same age because they take time off to raise children. However, Business Week reported on a study that compared the salaries of single white men and women between the ages of 20 and 40 (Koretz, 1990). When they factored out schooling, industry, skill level and work experience, the women still earned 91% of men's salaries. (Without factoring these out, women earned 86% of men's salaries.)

Another researcher analyzed the credentials of 194 corporate managers randomly chosen from 800 people who took a leadership course. He found that "if women were men with the same credentials, they would earn about 18 percent more" (Ridgon, 1993) And the 1992 edition of The Economics of Women, Men and Work found that less than 50 percent of the pay gap between men and women can be explained by differences in schooling and experience (Gries and Marsh, 1992). These figures are the closest estimate we have of the wage gap that can be explained only by discrimination.

References

Dwyer, Kelly, "IS Pay Dips; Women Still Lag," ComputerWorld, March 15, 1993, p. 1, 85-89.

Frenkel, Karen. "Women and Computing," Communications of the ACM, November, 1990, pp. 35-46.

Gries, David and Dorothy Marsh, "The 1989-90 Taulbee Survey," Communication of the ACM, January 1992, Vol 35, No. 1. pp. 133-143.

Koretz, Gene, "Economic Trends: Women Still Earn Less, But They've Come a Long Way," Business Week, December 24, 1990, p. 14.

Mahar, Maggie, "The Truth About Women's Pay," Working Women Magazine, April 1993, pp. 52-55, 100-103.

Pennar, Karen, "Women Are Still Paid the Wages of Discrimination," Business Week Oct 28, 1991, p. 35.

Rigdon, Joan, "Three Decades After the Equal Pay Act, Women's Wages Remain Far From Parity," Wall Street Journal, June 9, 1993, pp. B1, B10.

Schwartz, Joe, "Closing the Gap," American Demographics, January, 1988, pp. 10, 56.

Sims, Calvin, "The Unbreakable Glass Ceiling," The New York Times, June 7, 1993, pp. C1, C5.

Smith, Rebecca and James Mitchell, "A Stanford MBA Does Not Assure Equal Pay," San Jose Mercury News, Spring 1993

U.S. Department of Energy survey, "Degrees and Other Formal Awards Conferred," 1990-91.

U.S. Department of Energy survey, "Integrated Postsecondary Education Data System," 1990-91.

U.S. Department of Energy survey, "Completions," 1990-91.

White, Jane. A Few Good Women: Breaking the Barriers to Top Management. Prentice-Hall, 1992.