Goal & Concept
Goal
The goal of this exercise is to explore gender differences in occupation. Summary statistics, including the mean, median, mode, and standard deviation will be used.
Concept
The term sex is typically used to refer to a person's biological maleness or femaleness, whereas gender refers to the psychological, social, and cultural aspects of masculinity or femininity.
A large increase in women's labor force participation has occurred since World War II in the United States. Despite gains women have made, significant gender differences in occupational attainment remain. Women tend to be concentrated in office and administrative support and service occupations. In comparison, men tend to be concentrated in "blue collar" jobs, including skilled production, craft and repair work as well semi-skilled and unskilled manual jobs. Studies have found that women and men are equally likely to work in sales occupations.
There are two main theoretical explanations for the differences in occupational attainment between men and women. One explanation emphasizes the preferences and choices women make, which may be influenced by social pressures. The second explanation emphasizes gender discrimination in the labor market, which occurs when men and women with equal qualifications are treated differently.
Examples of possible research questions about gender and occupation:
- How are men and women distributed across occupations?
- Which occupations are predominantly female and which are predominantly male?
- How has the occupational distribution of men and women changed over time?
- Is the occupational distribution of men and women a result of gender discrimination or differences in the choices that men and women make?
Dataset
Data for this exercise come from a survey of employers which was one conducted as part of the Multi-City Study of Urban Inequality. The purpose of the study was to understand how changing labor market dynamics, racial attitudes and stereotypes, and racial residential segregation foster urban inequality.
The data represent a telephone survey of current business establishments in Atlanta, Boston, Detroit, and Los Angeles carried out between spring 1992 and spring 1995 to learn about hiring and vacancies, particularly for jobs requiring only a high school education. An employer size-weighted, stratified, probability sample (approximately two-thirds of the cases) was drawn from regional employment directories, and a probability sample (the other third of the cases) was drawn from the current or most recent employer reported by respondents to a household survey. Employers were asked about characteristics of their firms, including composition of the firm's labor force, vacant positions, the person most recently hired and his/her salary, hours worked per week, educational qualifications, promotions, the firm's recruiting and hiring methods, and demographic information for the respondent, job applicants, the firm's customers, and the firm's labor force, including age, education, race, and gender.
This exercise will use the following variables:
- Total number of clerical employees at firm (D2)
- Total number of female clerical employees at firm (D2A)
- Total number of sales employees at firm (D3)
- Total number of female sales employees at firm (D3A)
- Total number of manual or blue-collar employees at firm (D4)
- Total number of female manual or blue-collar employees at firm (D4A)
- On average, do you think that some of these tasks are performed better by men and others are performed better by women? (CC14)
Application
For this exercise, you will be exploring the distribution of men and women by occupation within firms using summary statistics.
First, to simplify analyses, we created new variables for all variables of interest to exclude "don't know/no opinion" (99998) and "refused/missing/skipped" (99999) answers. These variables are called D2new, D2Anew,...D4anew.
Next we created a variable, "ratiocler," for the proportion of clerical employees at a firm that is female by dividing the number of female clerical workers by the number of clerical workers overall. This variable ranges from 0 to 1, therefore a frequency distribution of ratiocler is a bit unwieldy.
Summary statistics provided below the table will help you to see the information more clearly. First look at the mean which represents the average proportion of a company's clerical workforce that is female. Next look at the value that divides the distribution in half (median) and the most frequently occurring value, the mode. Note the value of the standard deviation, which helps give a sense of how closely the values cluster around the mean. What can you say about these companies' clerical workforce -- does it tend to be more or less female than male?
Next we created a similar variable for the proportion of sales employees at the firm that is female, called "ratiosales."
Now examine the summary statistics for ratiosales. Again look below the frequency table at the summary statistics such as the mean, median, mode, and standard deviation. How does this distribution compare to the one you just saw for clerical workers?
The last created variable is similar to the first two and represents the proportion of blue-collar or manual employees at the firm that is female (Ratiomanual).
Now look at the summary statistics for ratiomanual. As before, examine the mean, median, mode, standard deviation, and the range of values. What do you find? What is the average proportion of female manual or blue-collar workers? What is the most common value among the firms in the sample (the mode)? Compare these statistics to those you had above. What can you say about the gender distributions of these companies?
Finally, given the discrepencies in the proportion of various job categories that are women, one might think that employers feel that men are just better at some tasks than women and vice versa. Is this the case? Look at the frequency distribution of answers to this question. What percent of respondents answered that they feel there are, on average, tasks for which men and women are better suited? Are the answers what you expected? Why or why not?
Interpretation & Summary
Think about your answers to the application questions before you click through to the interpretation guide for help in answering them.
What is the (arithmetic) average of the proportion of female clerical workers in these firms? What is the most commonly occuring value and the value that divides the distribution of companies in half? What can you say about these companies' clerical workforce -- does it tend to be more or less female than male?
What about for sales positions -- do companies have more or less equality in their gender distribution of sales positions?the median, mode, and standard deviation.
What do the mean, median, mode, and standard deviation tell you about the distribution of women in manual positions? Compare these statistics to those you had above. What can you say about the gender distributions of these companies?
What percent of respondents answered that they feel there are, on average, tasks for which men and women are better suited? Are the answers what you expected? Why or why not?
Interpretation
Things to think about in interpreting the results:
It is important to look at the amount of missing data in each table and think about the ways in which that might affect the generalizability of the results - some variables have relatively little missing data, others have a great deal. The ratio variables range from 32% missing data to 54% meaning that a significant number of those surveyed either didn't know or didn't answer one of the questions asking for the number of female or total employees in a particular job category.
Because the ratio variables can have any value between .00 and 1.00 (a continuous variable), the frequency distributions are difficult to interpret so we focus on the summary statistics, especially the mean, median, mode, and standard deviation. The mean tells the arithmetic average for any variable, it takes the proportion for each employer and adds them up, then divides by the number of employers answering the questions. The median is the value that, when the distribution is put into ascending/descending order, divides the distribution in half so that 50% of the employers are at or higher than that value and 50% are at or lower. The mode is simply the most frequently occuring value in the distribution -- the most employers gave that particular answer. The standard deviation is a measure of how closely (or not) values are clustered around the mean. The smaller the standard deviation, the closer each employer's answer was to the mean.
Proportions can be turned into percents by multiplying the value by 100.
The analyses show the following:
Companies have a much higher proportion of female clerical workers than male. In fact, over 50% of the employers reported having 100% female clerical workers with an overall average of 86% female.
Sales positions are more likely to be held by men than women. The mean of the distribution shows that just over 1/3 (35%) of sales employees were female, on average and 1/3 is the proportion above and below which 50% of the employers fell.
Manual positions are almost, but not quite, the inverse of clerical workers. That is, most companies reported having no female manual or blue-collar employees and the mean for the whole distribution is just over 1/4 (26%) female. The fact that the median is noticeably lower than the mean suggests that just a few companies reported having a higher proportion of female manual workers and those companies are pulling the mean upward (making things appear more equal than they would otherwise).
The standard deviations of the distributions show that the proportion of clerical workers is most tightly clustered around the mean and the proportion of manual workers is least clustered, though the three are not that different.
Even though distributions look highly unequal, employers overwhelmingly report that they do not feel that there are certain tasks at which men or women are just "better." Fully 71% denied feeling that men/women are better at certain tasks -- 83% if you look only at those who answered the question. This could be because employers really do not consciously feel this way, or because they felt uncomfortable reporting to the interviewer that they felt otherwise (e.g. the question could be affected by social desirability bias).
Summary
The goal of this exercise was to demonstrate gender breakdowns of various occupational categories. We examine three categories of jobs and find striking differences in the proportions of men and women performing each.
Bibliography
The references presented here represent resources that might be useful to instructors and students wishing to further explore this topic. All were chosen because they relate to the topic of study, whether or not they use the specific dataset that was used in this exercise. Some relate directly to the concepts as defined by the exercise, others explore the topic more broadly either conceptually or empirically. Most can be found in the ICPSR bibliography, though some outside sources were added if they were particularly relevant.
Albelda, Randy P., "Occupational Segregation by Race and Gender, 1958-1981." Industrial and Labor Relations Review. Apr 1986, 39, (3), 404-411.
Andes, Nancy, "Social Class and Gender: An Empirical Evaluation of Occupational Stratification." Gender and Society. Jun 1992, 6, (2), 231-251.
Bertrand, Marianne; Hallock, Kevin F. "The Gender Gap in Top Corporate Jobs." Industrial and Labor Relations Review. Oct 2001, 55, (1), 3-21.
Bridges, William P., "Rethinking Gender Segregation and Gender Inequality: Measures and Meanings." Demography. Aug 2003, 40, (3), 543-568.
Carrington, William J.; Troske, Kenneth R., "Gender segregation in small firms." Journal of Human Resources. Sum 1995, 30, (3), 503.
Cohen, Philip N.; Huffman, Matt L., "Occupational Segregation and the Devaluation of Women's Work across U.S. Labor Markets." Social Forces. Mar 2003, 81, (3), 881-908.
Cohen, Philip N.; Huffman, Matt L., "Individuals, Jobs, and Labor Markets: The Devaluation of Women's Work." American Sociological Review. Jun 2003, 68, (3), 443-463.
Correll, Shelley J., "Constraints into Preferences: Gender, Status, and Emerging Career Aspirations." American Sociological Review. Feb 2004, 69, (1), 93-113.
Garrison, Howard H., "Gender Differences in the Career Aspirations of Recent Cohorts of High School Seniors." Social Problems. Dec 1979, 27, (2), 170-185.
Goldin, Claudia. Understanding the Gender Gap: An Economic History of American Women. New York: Oxford University Press, 1990.
Jacobs, Jerry A., and Kathleen Gerson. The Time Divide: Work, Family, and Gender Inequality. Cambridge, MA: Harvard University Press. 2004.
Neumark, David. Sex Differences in Labor Markets. London: Routledge, 2004.
Perman, Lauri; Stevens, Beth, "Industrial segregation and the gender distribution of fringe benefits." Gender and Society. 1989, 3, (3), 388-404.
Petersen, Trond; Morgan, Laurie A., "Separate and Unequal: Occupation-Establishment Sex Segregation and the Gender Wage Gap." The American Journal of Sociology. Sep 1995, 101, (2), 329-265.
Reskin, Barbara F., and Heidi I. Hartmann. Women's Work, Men's Work: Sex Segregation on the Job. Washington, DC: National Academy Press, 1986.
Tienda, Marta; Smith, Shelley A.; Ortiz, Vilma, "Industrial restructuring, gender segregation, and sex differences in earnings." American Sociological Review. Apr 1987, 52, (2), 195-210.
Tomaskovic-Devey, Donald, "The Gender and Race Composition of Jobs and the Male/Female, White/Black Pay Gaps." Social Forces. Sep 1993, 72, (1), 45-76.
Walker, Henry A., "Gender Differences in Role Different and Organizational Task Performance." Annual Review of Sociology. 1986, 12, 255-275.
Warren, John R.; Sheridan, Jennifer T.; Hauser, Robert M., "Occupational Stratification across the Life Course: Evidence from the Wisconsin Longitudinal Study." American Sociological Review. Jun 2002, 67, (3), 432-455.
Watts, Martin, "Occupational Gender Segregation: Index Measurement and Econometric Modeling." Demography. Nov 1998, 35, (4), 489-496.
Weeden, Kim A., "Revisiting Occupational Sex Segregation in the United States, 1910-1990: Results from a Log-Linear Approach." Demography. Nov 1998, 35, (4), 475-487.
Wharton, Amy S., "Gender Segregation in Private-Sector, Public-Sector, and Self-Employed Occupations, 1950-1981." Social Science Quarterly. Dec 1989, 70, (4), 923-940.
Wharton, Amy S.; Baron, James N., "So Happy Together? The Impact of Gender Segregation on Men at Work." American Sociological Review. Oct 1987, 52, (5), 574-587.
Witkowski, Kristine M.; Leicht, Kevin T., "The Effects of Gender Segregation, Labor Force Participation, and Family Roles on the Earnings of Young Adult Workers." Work and Occupations. Feb 1995, 22, (1), 48-72.
Xu, Wu; Leffler, Ann, "Gender and Race Effects on Occupational Prestige, Segregation, and Earnings." Gender and Society. Sep 1992, 6, (3), 376-392.
