Sunteți pe pagina 1din 27

Salesmen, Saleswomen, or Sales Workers? Determinants of the Sex Composition of Sales Occupations Author(s): Marcia L.

Bellas and Barbara Thomas Coventry Reviewed work(s): Source: Sociological Forum, Vol. 16, No. 1 (Mar., 2001), pp. 73-98 Published by: Springer Stable URL: http://www.jstor.org/stable/685030 . Accessed: 09/11/2011 06:59
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

Springer is collaborating with JSTOR to digitize, preserve and extend access to Sociological Forum.

http://www.jstor.org

Sociological Forum, Vol. 16, No. 1, 2001

Salesmen, Saleswomen, or Sales Workers? Determinants of the Sex Composition of Sales Occupations'
Marcia L. Bellas2 and Barbara Thomas Coventry3 4

This study examines determinants of occupational sex composition in an expanding area of the economy-sales occupations. We used census data to determine which worker and occupational characteristicspredicted the percentage of women in sales occupations in 1980 and in 1990, and which variables contributed to changes in the percentage of women in sales occupations over time. Although sales occupations experienced a disproportionate decline in occupational sex segregation during the 1980s, we found that the same variables predicted the percentage of women in sales occupations in both 1980 and 1990. Average years of schooling, average hourly pay, the percentage of unemployed workers,and thepercentage of self-employed workers influenced the percentage of female workers in sales occupations in 1980 and in 1990, both for cross-sectional and dynamic models. In addition, a worker characteristic index that combines percentage White,percentage married, and average age of workers shows surprising effects on the percentage of women in sales occupations. Wediscuss the implications of thesefindings for occupational sex segregation, and suggest that a multidimensional queuing perspective offers the most useful interpretationof our results.
KEY WORDS: occupational sex composition; sex segregation; men and women in sales.

'Both authors contributed equally to the paper. We thank Chloe Bird, David Maume, and Sarah Smith for helpful comments on an earlier draft of this paper. 2Department of Sociology, P.O. Box 210378, 1018 Crosley Tower, University of Cincinnati, Cincinnati, Ohio 45221-0378. 3Department of Sociology and Anthropology, 2801 W. Bancroft, University of Toledo, Toledo, Ohio 43606; e-mail: barbara.coventry@utoledo.edo. 4To whom correspondence should be addressed. 73
0884-8971/01/0300-0073$19.50/0 ? 2001 Plenum Publishing Corporation

74

Bellas and Coventry

The American economy has changed dramatically during the second half of the twentieth century (Eitzen and Zinn, 1989; Kuttner, 1989). Manufacturing jobs, which enjoyed their greatest share of workers in the 1950s and 1960s, began to lose workers in the 1970s, while service sector jobs increased (Urquhart, 1984). Globalization of the economy, computer advances, and robotics have contributed to these changes (Dassbach, 1986; Eitzen and Zinn, 1989), along with women's increased labor force participation that helped spur the growth of service-sector occupations, such as sales (Macdonald and Sirianni, 1996). Indeed, by the year 2005, projections are that the manufacturing sector will employ only about 13% of workers, while the service sector will employ 81% (Reskin and Padavic, 1994:165-166). These economic shifts have contributed to changes in the distribution of workers across occupations. Concurrently, the sex composition of some occupations has also changed. Although most women continued to work in predominantly female occupations, a growing number were able to find work in predominantly male occupations, thus reducing occupational sex segregation during the 1970s and 1980s (Reskin and Roos, 1990). This paper examines sex segregation in one area of economic growthsales occupations. Aggregate sales occupations appear to have achieved integration. In 1990, women's overall representation in the 23 sales occupations recognized by the U.S. Census Bureau was only slightly greater than their share of the experienced civilian labor force-49 and 45 %, respectively (U.S. Bureau of the Census, 1993). However, aggregation often hides segregation. Examining detailed occupations instead of occupational groupings reveals that women and men are not similarly distributed. In 1990, men represented more than 85% of auctioneers, sales engineers, workers in motor vehicle and boat sales, and those in parts sales. Men accounted for more than 70% of sales representatives in mining, manufacturing, and wholesale; as well as workers selling hardware and building supplies; securities and financial services; and radios, televisions, hi-fis, and appliances. In contrast, women accounted for approximately 80% of cashiers; demonstrators, promoters, and models; and apparel salesworkers. In other sales occupations, such as real estate, furniture sales, and advertising and related sales, the proportion of women and men was approximately equal. Heterogeneity in the percentage of women in sales occupations is illustrated in Table I. Differences among sales occupations go beyond their sex composition, however. Because women and men are not evenly distributed across occupations, occupational rewards are not equally available to saleswomen and salesmen. In 1990, as Table I shows, cashiers (79% women) earned only $8,005 on average, while salesworkers in securities and financial services (27% women) averaged $53,812. These two occupations also differ markedly in prestige, working conditions, and job security (see U.S. Department

Determinants of the Sex Composition of Sales Occupations

75

Table I. Percentage of Women in Detailed Sales Occupations and Average Earnings in 1990 Average Percentage of women (%) earnings ($) Occupation (three-digit census codes) Sales engineers (258) Sales workers, parts (269-273) Sales workers, motor vehicles and boats (263) Auctioneers (284) Sales representatives, mining, manufacturing and wholesale (259-262) Sales workers, hardware and building supplies (268) Securities and financial services (255) Sales workers, radio, TV, hi-fi, and appliances (267) Supervisors and proprietors (243) Insurance sales (253) Other business services (257) News vendors (278-282) Sales workers, furniture and home furnishings (266) Real estate sales (254) Advertising and related sales (256) Sales workers, shoes (265) Sales counter clerks (275) Sales workers, other commodities (274) Street and door-to-door sales workers (277) Cashiers (276) Demonstrators, promoters and models, sales (283) Sales workers, apparel (264) 4.6 9.9 10.1 14.0 22.3 22.5 27.3 28.7 34.7 35.2 36.6 39.6 44.2 50.6 51.0 61.5 65.3 66.3 66.6 79.0 81.3 81.5 46,352 18,442 24,741 30,408 34,684 17,970 53,812 21,943 30,442 38,165 31,245 9,651 19,500 32,343 28,703 9,062 11,599 12,710 13,910 8,005 9,333 8,830

of Labor's Occupational Outlook Handbook for a description of these and other sales occupations). The strong association between the sex composition of occupations and occupational rewards-particularly earnings-attracted the attention of social scientists, who began monitoring changes in occupational segregation over time. Researchers found that for all occupations the index of segregation (Duncan and Duncan, 1955), the most common measure of occupational segregation, began to decline in the 1960s.5 The decline, primarily due to men's entry into predominantly female occupations, resulted in a segregation index of 67.6 in 1970. This figure indicates that 67.6% of men or women would have to change occupations to be evenly distributed across occupations. By 1980, the segregation index had declined to 59.8, largely due to women's movement into predominately male occupations (Jacobs, 1989). The index of segregation dropped even further, to 53, by 1990 (Reskin and Padavic, 1994).
5The formula for calculating the segregation index is: 1/2 Imit -fit 1, where mit = the percentage of the male labor force employed in the occupation at time t, and fit = the percentage of female labor force employed in the same occupation at time t.

76

Bellas and Coventry

This decline in occupational sex segregation was not uniform across occupational groups, however. The segregation index for sales occupations fell from 51.2 in 1980 to 39.9 in 1990, a decline of 11.3 points, considerably more than the 6.8 point decline for occupations overall (calculated from census data; U.S. Bureau of the Census, 1983, 1993). Our study examines determinants of the sex composition of sales occupations during this time of decreasing segregation. We use census data to address two questions: (1) What worker and occupational characteristics predict the percentage of women in sales occupations in 1980 and in 1990?, and (2) What variables contribute to changes in the percentage of women in sales occupations over time? We begin by outlining explanations of sex segregation, followed by a discussion of factors that may affect occupational sex composition. We next describe our data sources and methods, then discuss our findings and conclusions. EXPLANATIONS OF OCCUPATIONAL SEX SEGREGATION Patriarchy and Male Privilege Feminist labor market analysts claim that patriarchy,either alone or in conjunction with capitalism, is responsible for occupational sex segregation (Hartmann, 1981, 1976; Mitchell, 1974, 1971). The patriarchal system relegates women to low-paying, female-dominated jobs. Strober (1984, also see Strober and Arnold, 1987) contends that male workers, to whom employers give first choice of occupations and jobs, choose jobs that will maximize their economic gain. If a new occupation offers superior benefits, men move into it. If it compares unfavorably to existing occupational opportunities, men may refuse to enter it. Since employers can hire women at a lower wage rate than men, profit incentives should make women more attractive employees than men. However, according to Strober (1984), employers are also concerned with maintaining male privilege, since they themselves are typically men. Moreover, monetary costs include expenditures other than wages. Adherence to patriarchy can increase profits if the response of male workers and customers to female workers were to reduce productivity and sales, respectively. Work Place Culture, Sexual Harassment, and Discrimination Employer preferences for sales workers of one sex or the other may reflect their own preferences, or those of employees or customers. Moreover, these practices differ by occupation and may be justified on economic grounds. Blau (1984) argues that employers may have an occupation-specific objection to hiring women; for example, they will hire them in retail sales but

Determinants of the Sex Composition of Sales Occupations

77

not as car salesmen. Even when women are hired into predominately male jobs, gender stereotypes may manifest themselves in workplace cultures that exclude and isolate women, or more blatantly in sexual harassment and discrimination (Antilla, 1995; Catalyst, 1995; Kanter and Stein, 1979). These factors negatively affect women's work environments and opportunities, and may effectively drive women out of predominately male occupations (Jacobs, 1989). Employees may reinforce segregation by pressuring employers and managers to hire or promote one sex or the other for particular jobs. Hiring women into traditionally male jobs may threaten male workers' sense of masculinity, diminish their status, and raise fears that their earnings will decline (Williams, 1989). Male workers may respond by harassing women coworkers, or by creating hostile and alienating work environments. Employers may appease male workers to avoid disruptions in the workplace and possible reductions in worker productivity, or out of loyalty to same-sex employees (Cockburn, 1988; Reskin and Roos, 1990:309-310). Customers may also prefer salesworkers of one sex or the other, particularly when the position requires technical knowledge or expertise about the products. In insurance sales, for example, "some men do not like to discuss financial affairs with women. They have a feeling that no woman can know as much as a man might" (Karene and Anderson, 1967:5; reported in Thomas, 1990:189). In her study of car saleswomen, Lawson, (1993:11) found that some customers responded to women salespersons by simply refusing to deal with them ("Get me a real salesman! What the hell are you doing here?"), or by asking saleswomen questions that they would not ask a salesman (e.g., "How much does the engine weigh?"). Customers may prefer saleswomen in certain contexts, of course, assuming for example that women are more knowledgeable than men about clothing or housewares (Game and Pringle, 1983). Because women constitute a growing proportion of customers in some areas (e.g., automobiles, insurance, and real estate) and some salesmen don't want to sell to women, employers may hire women to sell in expanding "women's markets." As a rule, however, these markets tend to be associated with lower earnings than others (Thomas, 1990). In real estate sales, women sell homes but are scarce in the more lucrative area of commercial sales. Men also dominate managerial positions (broker-manager), while the sales staff is likely to comprise more women (Thomas and Reskin, 1990). In car sales, managers may place women near less expensive cars and instruct them to allow salesmen to handle customers interested in more expensive cars (Lawson, 1993). In some instances, there is a clear association between the sex of the salesperson, the gendered nature of products, and the sex composition of clientele (Berheide, 1988; Bradley, 1989; Game and Pringle,

78

Bellas and Coventry

1983, Chapter 3). However, in most cases, the underlying principle seems to be simply that the most lucrative positions are by and large reserved for men. Employers serve as gatekeepers in the hiring process, and male employees use various mechanisms to ward off female interlopers. Human Capital Theory Human capital theory proposes that women voluntarily choose to enter lower-paying occupations because these occupations require lower investments in human capital, and because time spent out of the labor force is accompanied by lower wage depreciation (Polachek, 1979). Thus women's occupational distribution presumably results from rational decisions to limit investments in human capital and reduce any deleterious consequences of interrupted labor force participation. Consistent with the human capital perspective, some women may choose to limit the hours they work or select jobs that require little or no travel, or offer self-employment opportunities. Women in real estate and insurance sales frequently cite flexibility, independence, and part-time hours as reasons for pursuing these lines of work. These sales occupations may offer less of these amenities than is commonly believed, however, particularly in the difficult first years of building a clientele (Thomas, 1990;Thomas and Reskin, 1990). Women may also seek occupations with more self-employment opportunities to better balance work and family demands (Biggart, 1996;Loscocco, et al., 1991). Although concerns about balancing work and family demands may influence some women's occupational choices, research provides limited support for the human capital explanation of occupational sex segregation. Predominately female jobs are not especially accommodating of family demands (Glass, 1990; Jacobs and Steinberg, 1990), women are no less committed to their jobs than men, on average (Marsden et al., 1993), and women and men do not differ in their energy expenditures on the job (Bielby and Bielby, 1988). Despite such evidence to the contrary, employers' beliefs that women are less committed or productive workers than men can influence hiring decisions and therefore contribute to sex segregation (Arrow, 1972). Gender-Role Socialization Human capital theory assumes that a woman's primary concern is her family, not a job or a career. Gender-role socialization theories suggest that women and men are taught to aspire to gender-typed occupations. In other words, girls and boys learn which occupations society deems most

Determinants of the Sex Composition of Sales Occupations

79

appropriate for their gender, and learn skills relevant to these occupations. Children may also learn to prefer to spend time with members of their own sex, which may also contribute to occupational segregation (Reskin and Padavic, 1994). As is the case for human capital arguments, there is little empirical support for the view that socialization is an important cause of occupational sex segregation. Jacobs (1989) found almost no relationship between the occupational aspirations of youth and the occupations they held as adults. Furthermore he found that men, but to a greater extent, women, circulate between predominately male and predominately female occupations. Segmentation Theories Other analysts contend that occupational sex segregation results from the confinement of different types of workers to different labor markets (Edwards, 1979; Gordon, 1972; Sokoloff, 1980). Segregation theorists often identify two markets: the secondary labor market, characterized by low wages, high job instability, poor working conditions, and few benefits, and the primary labor market, which provides high wages, job stability, good working conditions, and benefits. Gordon (1972) describes the secondary labor market as being dominated by people of color, women, and teenagers, while adult White men are concentrated in the primary market. Although "dual labor market analysts refer to the employment problems of women only implicitly" (Gordon, 1972:48), those interested in sex differences in labor market outcomes have argued that men and women operate in separate markets because men and women do not tend to compete with each other for the same jobs (Reskin and Roos, 1990; Sokoloff, 1980). Moreover, on the limited occasions when women are able to secure employment in the primary market, they often find themselves segregated within that market (Blau and Jusenius, 1976; Sokoloff, 1992). Although segmentation theories offer an alternative to the singlemarket perspective, other analysts have criticized segmentation theories (Milkman, 1987; Sokoloff, 1980). Critics argue that segmentation theories examine inequality between men and women, but they are basically descriptive and neglect to explain how men and women were segregated initially into different markets. They are also characterized as being too rigid and simplistic. Although Edwards (1979) divides the primary market into "independent primary"and "subordinate primary"segments, this approach is still inadequate because it continues to treat women and other pools of cheap labor similarly (Milkman, 1987). Even though, traditionally, segmentation theorists have tended to deal with women only in passing (Gordon, 1972),

80

Bellas and Coventry

those who focused on women's experiences also come to the conclusion that multiple markets do exist, and that men and women operate in different markets (Reskin and Roos, 1990; Sokoloff, 1980). Queuing Theory Queuing theory argues that workers are segregated because employers rank workers (and workers rank jobs) according to their desirability. Because employers lack direct measures of potential workers' future productivity, they use educational attainment, work experience and ascribed characteristics (e.g., sex, race, ethnicity) as proxies to screen job applicants (Lieberson, 1980; Thurow, 1969, 1975). When employers cannot attract the most preferred employees, they move down the queue to hire lower-ranked workers (Reskin and Roos, 1990; Thurow, 1969, 1975). If employers prefer to hire men over women, men should occupy the most highly ranked jobs, leaving lower-queued jobs for women (Reskin and Roos, 1990). Although demographic characteristics influence workers' placement in a labor queue, they are not the only ranking criteria. Overlap can exist between male and female labor queues, especially when employers have little aversion to hiring women; in this case, the top-rated women are ranked higher than the lowest-rated men. While men as a group have first choice of jobs, not all men will occupy the best positions in the job queue because they do not have the requisite qualifications and/or skills (Reskin and Roos, 1990). In addition, employers of labor-intensive jobs may prefer women over men in order to reduce labor costs (Cohn, 1985). However, in these cases, Reskin (1988) argues that it is lower-status employers (e.g., those that pay less well) who rank women over men, so this does not threaten the patriarchal system but in fact supports it. PREDICTORS OF OCCUPATIONAL SEX COMPOSITION These explanations of occupational segregation suggest factors that might affect the sex composition of sales occupations. Labor market analysts indicate that sex segregation might be associated with other demographic characteristics as well. Gordon (1972) point outs that race and age, in addition to sex, are related to workers' placement in labor markets, since people of color, women, and teenagers are overrepresented in the secondary labor market. Employers might also consider a potential workers' marital status when making employment decisions. Kanter and Stein's (1979) study of women in industrial sales found that most complaints by women centered

Determinants of the Sex Composition of Sales Occupations

81

around stereotypes held by employers or bosses about marriage and childbearing. While being married may hurt women, marriage appears to help men achieve career success (Bellas, 1992; Finch, 1983; Fowlkes, 1980). Consequently, married men may tend to concentrate in more desirable occupations, and therefore those with fewer women. Marital status undoubtedly correlates with age, younger workers being least likely to be married. Thus, consistent with Gordon's (1972) perspective of labor market segmentation, one would expect women to be concentrated in sales occupations with more nonWhite, younger, and unmarried workers. However, Milkman (1987) indicates that employers segregate women from other sources of cheap labor, suggesting that women tend to be concentrated in different sales occupations than nonWhite, younger, and unmarried workers. In addition to sex, race, age, and marital status, queuing theorists suggest that employers may use educational credentials to screen applicants (Lieberson, 1980; Thurow, 1969; for sales occupations also see U.S. Department of Labor 1996). Educational prerequisites may affect some women's entry into certain occupations. Although the percentages of bachelor's and master's degrees earned by women now exceed those of men, educational requirements may influence the occupational choices of older women who reenter the labor market. In real estate, for example, relatively low educational requirements as well as the initiation of more formal entry mechanisms (e.g., community college courses) rather than previous broker sponsorship have contributed to women's increased representation in the occupation (Thomas and Reskin, 1990).6 Consequently, we expect women to be concentrated in sales occupations whose workforces have lower average years of education. Labor market analysts offer competing views about the relationship between earnings and occupational sex segregation. According to feminist and queuing theorists, employers who are unable to attract male workers at prevailing wage levels (or who are unwilling or unable to pay higher wages) may turn to women to fill their labor needs (Reskin and Roos, 1990; Strober, 1984). Human capital theory, on the other hand, contends that women are concentrated in lower-paying jobs than men because women choose to invest less in their human capital. Based on both of these competing explanations, we expect women to be concentrated in sales occupations with lower average wages.
6Sex segregation in college majors may also contribute to occupational sex segregation. Because men are more likely than women to major in science and engineering (Ransom, 1990; Jacobs, 1985), men may account for a larger percentage of the applicant pool for jobs requiring technical expertise. Yet, many occupations, including most sales occupations, do not require specialized backgrounds even when employers prefer college-educated workers (U.S. Department of Labor 1996). Moreover, workers with the same college major are segregated by sex.

82

Bellas and Coventry

Other occupational characteristics besides wage levels and educational attainment may affect employers' demand for male and/or female salesworkers. These occupational characteristics include union membership, unemployment rates, occupational growth, and self-employment. Historically, labor unions have served "gatekeeping" functions by excluding women (and people of color) (Amott, 1995). Although unions no longer formally bar these groups, Benson (1984) notes that unions have often not dealt with issues of interest to women, thereby discouraging their membership. Consequently, we anticipate that women will be most heavily represented in sales occupations with lower rates of unionization. However, unionization may represent less of a barrier to women in sales than to those in some other occupations since, with the exception of grocery store clerks, sales occupations have low unionization rates (Hartmann et al., 1987; Benson, 1984). In eight of the fourteen sales occupations analyzed by Kokkelenberg and Sockell (1985), the proportion of unionized workers in 1980 was less than 2%; in 13 of 14 sales occupations, unionization was less than 6%. The unionization rate for the labor market overall was about 20% in 1980 (U.S. Bureau of the Census 1996:438, Table 684). Occupational "decline" may result in men being less willing to pursue jobs in an occupation, thereby creating opportunities for women (Reskin and Roos, 1990). Higher unemployment rates for men may reflect a decline in the number of job opportunities overall or indicate that women have replaced some men in these occupations, perhaps because employers sought cheaper labor. In addition to examining the effect of unemployment rates on occupational sex composition, we attempt to capture some of these dynamics by exploring the effects of occupational growth for men and women. We expect that increased employment opportunities for men in sales occupations will also result in greater opportunities for women, provided a shortage of men ensues (Reskin and Roos, 1990; also see Sokoloff, 1992).7 Therefore, we expect that male occupational growth will have a negative effect on the percentage of women in occupations, while female occupational growth will have a positive effect. Existing literature presents conflicting hypotheses regarding the effect of self-employment opportunities on occupational sex composition. According to human capital theory, women may seek occupations with more self-employment opportunities to better balance work and family demands (Biggart, 1996; Loscocco et al., 1991). Self-employment might also attract women who wish to avoid discriminatory employers. However, women may
7Note that an influx of women into a predominately male occupation (or vice versa) does not necessarily result in an integrated occupation. Rather, it may reflect a process of resegregation (Bradley, 1993; Reskin and Roos, 1990).

Determinants of the Sex Composition of Sales Occupations

83

have fewer self-employment opportunities than men because women may be discriminated against in business loans or other aspects of starting their own business (Cotter et al., 1995). Because self-employment rates for women have increased faster in recent years than rates for men, we hypothesize that women will be more highly concentrated in sales occupations with higher self-employment rates. Note, though, that self-employment is still less prevalent among women, and female-owned businesses generate far less sales revenue than male-owned businesses (Amott, 1995; Loscocco et al., 1991). DATA AND METHODS Our interest is in determining what worker and occupational characteristics influenced the percentage of women in sales occupations in 1980 and 1990, and which variables contributed to changes in the percentage of women in sales occupations over time. The independent variables in our analyses include many of those suggested in the literature as influencing the percentage of women in sales and other occupations.8 Data Unless indicated, we constructed aggregate-level variables for sales occupations using the U.S. Census Bureau's Public-Use Micro Sample data for 1980 and 1990 (U.S. Bureau of the Census 1983, 1993). We used data for the civilian labor force to calculate worker and occupational characteristics for 22 sales occupations (three-digit occupational codes, 243-284).9 The civilian labor force includes both employed and unemployed workers aged 16 years and older. Measures Following Treiman and Terrell (1975), we constructed aggregate-level measures using average characteristics of male workers rather than those of all workers for all variables except unionization rates and occupational growth. This method avoids confounding the independent variables with the
8Although licensing may represent a barrier to women's employment in some occupations we excluded licensing because only three sales occupations require licenses, and the regression coefficient was not statistically significant. 9We omitted "sales occupations not elsewhere classified" since it is a catch-all category that has little interpretive value.

84

Bellas and Coventry

dependent variable, the percentage of women in sales occupations (see also, Reskin and Roos, 1990:44-46).10 For example, if women concentrate in occupations with higher unemployment rates for men, this method eliminates the possibility that a relationship between unemployment and percentage female simply reflects the aggregation of unemployment rates for women. Dependent Variable Percentagefemale is the percentage of women in each sales occupation. Independent Variables (Characteristicsof Men Employed in Each Sales Occupation) We constructed an index of worker characteristicsby averaging z-scores for three variables-percentage White (percentage of male workers in each sales occupation who are non-Hispanic Whites), percentage married (percentage of male workers in each sales occupation who are married), and age (average age of male workers in each sales occupation). Higher-index scores reflect those workers who, according to structural explanations, employers would prefer to hire if they could attract them (Whites, married and relatively older workers).l1 The alpha reliability for the worker characteristics
index is .936 for 1980 and .961 for 1990.12

Independent Variables (Characteristicsof Sales Occupations) Years of schooling is the average for male workers in each sales occupation. Hourly pay is the average for male workers in each sales occupation. We constructed this variable by first multiplying the number of weeks employed
10Treimanand Terrell (1975) used characteristics of White male workers for some, but not all, variables. Since we included the percentage of White male workers as a variable in the analysis, we elected to construct variables using both White and nonWhite male workers combined. 11Some employers may view older workers as less desirable than younger workers because wages are higher for more experienced workers or because they fear older workers' skills are obsolete. Because Gordon (1972) contends that women and teenagers share the secondary labor market and the correlation between percentage female and age is negative, we anticipated that women concentrated in occupations with younger workers more so than in occupations with older workers. We constructed the worker characteristics index to reflect this. 12Thereliabilities would increase marginally if we deleted percentage white from the 1980 index and average age from the 1990 index. However, we retained the items to provide consistent measures for 1980 and 1990.

Determinants of the Sex Composition of Sales Occupations

85

in 1979 (for the 1980 analysis), and 1989 (for the 1990 analysis) by the usual number of hours employed each week. We divided the product into 1979 and 1989 income, respectively (sums wages or salary income plus any nonfarm, self-employment income). Percent unemployed is the percentage of male workers in each sales occupation who were not employed and were looking for work. Percent self-employed is the percentage of male workers in each sales occupation employed in their own incorporated or unincorporated business. Percent unionized represents the percentage of workers in each sales occupation who belonged to a labor union in 1980, the most recent year available (Filer, 1988). Because separate unionization rates for male and female workers are not available, the measure is based on the overall occupational workforce. Occupational growth refers to change in the size of the civilian labor force in each sales occupation between 1970 and 1980, and between 1980 and 1990. We constructed two growth measures, one for males and one for females. We first determined the difference between the number of same-sex workers in 1970 and 1980, and 1980 and 1990, then for each decade calculated the difference as a percentage of the base year figure. Table II displays values for the variables included in the 1990 analysis for each of the 22 sales occupations, as well as their means and standard deviations. Analysis After examining descriptive statistics and correlations for the variables, we constructed two cross-sectional regression models to predict the percentage of women in sales occupations in 1980 and in 1990. The models include average worker and occupational characteristics in 1980 and 1990, respectively. We included the percentage of unionized workers in both models, although the measure reflects unionization rates for 1980. We could not obtain later unionization rates for detailed sales occupations and thought including somewhat dated figures in the 1990 model were preferable to excluding the variable altogether.13 Because of the likelihood of heteroscedasticity in aggregate-level data, we used estimated generalized least squares (EGLS) regression. We weighted each variable by the square root of the number of men in each sales occupation in 1980 (and 1990), and included the square root as an independent
13Unionization rates were probably lower in 1990 than in 1980 (although the two rates are undoubtedly highly correlated), as unionization among U.S. workers has been declining steadily. This decline has been slower in service occupations than in other economic sectors, however (U.S. Bureau of the Census 1996:438, Table 684).

Table II. Average Characteristics of Salesworkers and Sales Occupations

Occupation Sales engineers Parts Motor vehicles and boats Auctioneers Mining, manufacturing, wholesale Hardware and building supplies Securities and financial services Radio, TV, hi-fi, and appliances Supervisors and proprietors Insurance Other business services News vendors Furniture and home furnishings Real estate Advertising and related sales Shoes Counter clerks Other commodities Street and door-to-door Cashiers Demonstrators, promoters and models Apparel Mean Standard deviation

Se Averag;e Female White Married Average years Unionized Unemployed emp ofscho ol (%) (%) (%) (%) age (%) (% 2.2 4.6 78.1 41.9 15.2 1.9 95.7 7 56.7 12.5 3.5 9.9 89.1 34.6 3.3 5 10.1 89.6 65.0 13.1 2.9 39.7 4.7 13 12.8 14.0 96.3 82.0 45.9 .0 2.6 64 22.3 40.9 14.1 2.9 2.4 92.8 73.4 14 22.5 27.3 28.7 34.7 35.2 36.6 39.6 44.2 50.6 51.0 61.5 65.3 66.3 66.6 79.0 81.3 81.5 42.4 24.2 97.3 92.5 87.4 88.5 90.9 91.7 83.9 89.1 90.8 89.4 71.8 81.3 84.7 84.0 73.5 83.5 76.2 87.0 6.6 58.9 68.0 46.7 73.8 79.2 66.5 39.1 61.6 71.7 60.0 28.4 43.4 46.5 47.2 26.9 44.8 33.9 56.9 16.7 37.9 39.7 33.2 40.0 43.7 39.3 33.9 40.8 46.1 39.0 30.6 33.5 36.1 36.4 29.4 38.6 32.5 37.9 4.6 13.1 15.6 14.0 13.7 14.8 14.3 11.7 13.3 14.8 14.6 12.8 12.9 13.3 13.5 12.4 13.2 13.1 13.6 1.0 4.4 .7 3.3 5.4 5.6 10.2 .9 3.2 3.1 5.8 5.7 12.9 5.9 2.0 19.1 .6 5.7 4.8 4.4 3.2 2.9 5.6 2.4 2.4 1.9 4.0 8.3 3.8 1.9 4.9 8.6 6.7 5.9 9.5 9.7 15.6 5.4 3.5

20

10

23 32 15 20 18

45. 16

5 17 18 17 4. 16

10 18 13

Determinants of the Sex Composition of Sales Occupations

87

variable in the regression equation. The coefficient for the square root is the constant. This procedure gives more weight to aggregate-level measures based on larger numbers of cases since they are presumably more accurate and should show less variance than measures based on smaller numbers of cases (Hanushek and Jackson, 1977:142-153). We then constructed a dynamic model to assess the effects of the independent variables for 1980 on the percentage of women in sales occupations in 1990. In this model, we weighted each variable by the square root of the number of men in each occupation averaged for 1980 and 1990. With a dynamic model we can determine the extent to which the independent variables contributed to changes in the percentage of women in sales occupations between 1980 and 1990. By including the percentage of women in 1980 as an independent variable, this model also holds constant any independent variable omitted from the regression model, provided they are correlated with both 1980 and 1990 percentage female (Hannan and Young, 1977). RESULTS 1980 Cross-Sectional Analysis We found that all three characteristics of male workers-percentage White, percentage married, and average age-were significantly correlated with the percentage of women in sales occupations in 1980 (data not shown). Women tended to concentrate in occupations with more nonWhite, unmarried, and younger male workers. Four of the six occupational variables were significantly correlated with the percentage of women in sales occupations. Occupations with higher percentages of women tended to have higher rates of unemployment and unionization, and lower pay. In addition, average years of schooling was marginally significant in its negative association with percentage female (p < .10), indicating that occupations with higher proportions of women were characterized by lower educational levels among male workers. These relationships are consistent with general segregation patterns in the larger labor market. We next sought to determine which independent variables retained their relationships with the dependent variable after controlling for the effects of other variables. The results of regressing the percentage of women in sales occupations in 1980 on worker and occupational characteristics are shown in Table III. Step 1 of the 1980 regression shows that the worker characteristics index exerted a significant negative effect on percentage female. Thus, controlling for the effects of education, wage level, unionization, unemployment, self-employment, and male occupational growth, women were most

88

Bellas and Coventry

Table III. Regression of the Percentage of Women in Sales Occupations in 1980 and 1990 on Average Male Worker and Occupational Characteristics 1980 Step 1, b (SE of b) Square root of N (intercept) Index of worker characteristics Average years of schooling Average hourly pay Percent unionized Percent unemployed Percent self-employed Male occupational growth (percentage change, past 10 years) Female occupational growth (percentage change, past 10 years) R2 Step 2, b (SE of b) Ste 1, b zp (SIEof b) 1990 Step 2, b (SE of b)

-378.93 -391.27 -22z4.32 -235.03 (90.77) (91.63) (8.9.19) (118.08) -26.69*** -20.66** -1: 3.94** -13.44* (9.18) (11.00) 7.42) (8.42) 31.60**** 32.26**** (1 9.84** 20.72** (7.56) (7.60) 7.65) (9.96) -4.51** -4.60** -2 3.00** -3.22* (2.21) (2.21) (c 1.38) (2.06) .35 .75 .80* .86 (1.21) (1.28) (.54) (.71) 7.33** 8.90** 3.43** 3.64* (4.13) (4.42) (c 1.79) (2.34) 1.01**** .95*** 1.23**** 1.26*** (.26) (.27) (.23) (.29) -.02 -.02E-1 .01 -.03E-1 (.11) (.11) (.04) (.11) -.02 .01 (.02) (.08) .37 .40 .78 .78

Note. N = 22 (13 degrees of freedom). *p < .10; **p < .05; ***p < .01; ****p < .001 (one-tailed tests).

prevalent in occupations with more nonWhite, younger, and unmarried male workers. Adding female occupational growth to the model (Step 2) does not substantively change the results, but it does increase the explained variance to 40%. As Step 2 shows, each unit decrease in the worker characteristics index is associated with a 20% increase in women within an occupation. Table III also shows that average years of schooling strongly influenced the percentage of women in occupations in 1980, though not in the direction suggested by the bivariate relationship between these variables. All else equal, women were more prevalent in sales occupations with higher educational requirements than those with lower ones. While this finding is consistent with the interpretation that women have moved into occupations requiring higher educational levels (Cotter et al., 1995; Jacobs, 1992), it does not mean that workers are compensated for their higher educational levels. Because the regression controls for wage level, the independent effect of education indicates that salesmen (and saleswomen, since the correlation between their educational levels is .949) in occupations with higher proportions of women were more highly educated than comparably paid workers in occupations with fewer women. In other words, workers in occupations with more women appear to have been undercompensated for their education.

Determinants of the Sex Composition of Sales Occupations

89

This is consistent with a comparable worth perspective, which argues that occupations, jobs, and job tasks that are associated with devalued groups (e.g., women), are compensated less than comparable occupations, jobs and job tasks performed by more highly valued groups (e.g., men) (England, 1992; Steinberg, 1990). Average hourly wage also exerts a direct negative effect on the percentage of women in sales occupations. Not only were those salesworkers in occupations with higher proportions of women not fully compensated for their educational levels in 1980, but they also suffered an independent wage penalty. The effect of the percentage of unemployed male workers on percentage female also achieved statistical significance, indicating that sales occupations with higher unemployment rates for men tended to have more women. This finding is consistent with a queuing interpretation. At times, men eschew certain occupations or jobs (e.g., because of declining wages or working conditions); this may create opportunities for women (Reskin and Roos, 1990). Note, however, that neither occupational growth variable directly influenced the percentage of women in sales occupations. The percentage of self-employed male workers directly influenced the percentage of women in sales occupations in 1980. Sales occupations with more self-employed men tended to include more women, controlling for the effects of other variables. Yet the intersex correlation between men and women's self-employment rates is negative, indicating that occupations with more self-employed men had fewer self-employed women. This finding may reflect sex segregation within occupations, with men more likely to be selfemployed proprietors and women more likely to work for others.14 1990 Cross-Sectional Analysis To assess stability in the variables that influence the percentage of women in sales occupations, we repeated the regression, using 1990 data. The correlations between the percentage of women in sales occupations in 1990, and the worker and occupational characteristics were similar to those for 1980 (data not shown), except that male occupational growth was positively and significantly associated with percentage female in 1990. Occupations experiencing growth for men had higher proportions of women.
14Thisraises the question of whether constructing variables using characteristics for men misrepresents the situation for women workers (Treiman and Terrell, 1975). With few exceptions (the correlations between the percentages of married male and female workers and male and female occupational growth from 1980 to 1990) the characteristics of female and male workers are highly correlated. Of these correlations, all are positive except that between the percentages of self-employed male and female workers.

90

Bellas and Coventry

In addition, years of schooling was no longer significantly correlated with percentage female, and the correlation between average hourly pay and percentage female became less significant (p < .10). Table III shows the results of regressing the percentage of women in sales occupations in 1990 on worker and occupational characteristics. The findings for 1980 and 1990 are quite similar, though, interestingly, the 1990 model explains twice as much variance as the 1980 model. As Step 1 of the 1990 model shows, the worker characteristics index exerts a negative effect on the percentage of women in sales occupations in 1990. Adding female occupational growth in Step 2 causes the index to become marginally significant in its negative effect on percentage female (p < .10). Consistent with the 1980 results, average years of schooling and percentage self-employed positively affected percentage female in 1990. Average hourly pay negatively and percentage unemployed positively influenced percentage female, though these effects were only marginally significant (p < .10). Dynamic Model We used a dynamic model to assess which variables appear to influence changes in the percentage of women in sales occupations between 1980 and 1990. Because a dynamic model can determine whether any of the independent variables for 1980 contribute to change in the dependent variable, it is a more powerful causal model than traditional cross-sectional analyses. As previously indicated, the index of segregation for sales occupations fell from 51.2 in 1980 to 39.9 in 1990, a decline of 11.3 percentage points (U.S. Bureau of the Census, 1983, 1993). Table IV shows the percentage of women in each of the 22 sales occupations in 1980 and 1990, as well as the percentage change over the 10-year period. Insurance sales experienced the greatest increase in women between 1980 and 1990 (+10.2%), while door-to-door sales experienced the greatest decline (-11.1%). To put these figures in perspective, between 1980 and 1990, the increase in women's overall labor force participation was 5.9% (Blau and Ferber, 1992:75). Therefore any change greater than 5.9% represents a disproportionate increase in the percentage of women workers; anything less represents a disproportionate decrease. Since most sales occupations had substantial over- or underrepresentations of women in 1980, changes between 1980 and 1990 should be considered in the context of women's initial representations. As Table IV illustrates, most of the twenty-two sales occupations experienced either a decline in the percentage of women between 1980 and 1990 or a rate of increase less than that for the labor force overall. Table V shows the results of the dynamic regression model. In Step 1, we entered all independent variables except female occupational growth and

Determinants of the Sex Composition of Sales Occupations

91

Table IV. Percentage of Women in Sales Occupations in 1980 and 1990, and Percentage Change Between These Years Percentage of females 1980 1990 Percentage change 3.5 7.1 7.8 14.3 14.4 18.3 24.1 25.0 27.4 27.6 33.0 36.6 40.4 41.7 44.6 56.4 72.1 72.1 76.9 77.6 81.0 83.0 4.6 10.1 9.9 22.3 14.0 27.3 22.5 35.2 34.7 28.7 39.6 36.6 44.2 51.0 50.6 61.5 66.3 65.3 81.3 66.6 81.5 79.0 +1.1 +3.0 +2.1 +8.0 -.4 +8.0 -1.6 +10.2 +7.3 +1.1 +6.6 0.0 +3.8 +9.8 +6.0 +5.1 -5.8 -6.8 +4.4 -11.1 +.5 -4.0

Occupation Sales engineers Sales workers, motor vehicles and boats Sales workers, parts Sales representatives, mining, manufacturing and wholesale Auctioneers Securities and financial services Sales workers, hardware and building supplies Insurance sales Supervisors and proprietors Sales workers, radio, TV, hi-fi, and appliances News vendors Other business services Sales workers, furniture and home furnishings Advertising and related sales Real estate sales Sales workers, shoes Sales workers, other commodities Sales counter clerks Demonstrators, promoters and models, sales Street and door-to-door sales workers Sales workers, apparel Cashiers

the percentage of women in 1980, which we added in Steps 2 and 3, respectively. The results closely mirror those for the 1980 model, indicating that factors that predict the percentage of women in sales occupations in 1980 also predict changes in the percentage of women over time. Controlling for the effects of all variables, except percentage female in 1980, the worker characteristics index for 1980 negatively affects the percentage of women in sales occupations in 1990 (Step 2). Each unit decrease in the 1980 worker characteristics index is associated with approximately 19% more women in an occupation in 1990. Thus, sales occupations showing the slowest increase in the percentage of women workers between 1980 and 1990 were those with higher percentages of White, married, and relatively older (male) workers in 1980. In addition, average years of schooling and percentage of self-employed workers in 1980 positively influenced the percentage of women in sales occupations 10 years later. The percentage of unemployed workers and hourly pay in 1980 negatively affected the percentage of women in occupations in 1990. After controlling for the percentage of women in sales occupations in 1980 (Step 3), the effects of other independent variables disappeared, except for a marginally significant effect of unionization.

92

Bellas and Coventry

Table V. Regression of the Percentage of Women in Sales Occupations in 1990 on Average Male Worker and Occupational Characteristics in 1980 Step 1, b (SE of b) Square root of N (intercept) Index of worker characteristics Average years of schooling Average hourly pay Percent unionized Percent unemployed Percent self-employed Male occupational growth Female occupational growth Percent female in 1980 R2 .46 -282.65 (83.95) -25.18*** (7.99) 23.71*** (6.83) -2.43 (1.80) .88 (.92) 4.74 (3.76) .98**** (.24) -.04 (.09) Step 2, b (SE of b) -297.79 (83.95) -18.94** (9.55) 24.61*** (6.79) -2.55* (1.79) 1.23 (.95) 6.39* (3.98) .92*** (.24) -.02 (.09) -.02 (.02) .49 Step 3, b (SE of b) .86 (51.07) -2.34 (4.47) .08 4.17 .74 (.85) .71* (.40) -.63 (1.87) .21 (.13) -.01 (.04) -.01 (.01) .79**** (.10) .93

Note. N = 22 (12 degrees of freedom). *p < .10; **p < .05; ***p < .01; ****p < .001 (one-tailed tests).

As one would expect, all else being equal, sales occupations with the highest percentages of women in 1980 were those with the highest percentages of women in 1990. DISCUSSION AND CONCLUSIONS In constructing our regression models, we drew on a variety of explanations of occupational sex segregation. While offering conflicting expectations for some predictors, these diverse approaches often suggest similar effects of predictor variables on the percentage of women in sales occupations. For example, patriarchy,segmentation, human capital and queuing perspectives all suggest that women will be concentrated in sales occupations with lower average years of education and lower wage levels. Although our results show the expected negative effect of wages on the percentage of women in sales occupations, the relationship between education and occupational sex composition appears more complicated. The positive effect of education on percentage female is consistent with research suggesting that rising educational levels for women have provided increased entry into some of the better male-dominated occupations (Cotter et al., 1995; Jacobs, 1992).

Determinants of the Sex Composition of Sales Occupations

93

Our analyses also revealed that women were more prevalent in sales occupations with higher educational levels after controlling for other factors, including wage level. Thus, workers in sales occupations where women predominate appear to be penalized financially relative to those in occupations where women are less prevalent. While this finding supports various conflict approaches (e.g., patriarchy,segmentation theory, queuing theory), it is contrary to the tenets of human capital theory which claim that women earn less than men because they invest less in education and other forms of human capital. Our results also show that the greater the percentage of self-employed men in sales occupations, the more women in these occupations. The positive effect of self-employment on percentage female appears to support the human capital view that women may choose self-employment in order to balance family and work. A more conflict-oriented approach might suggest that women's motives for choosing self-employment could also include escape from discriminatory employers. Regardless of the reason for women's decisions, the correlation between the self-employment rates for men and women is negative. That is, higher self-employment rates for men do not correspond to higher self-employment rates for women. We interpret this finding as indicative of a finer level of vertical sex segregation within occupations than we can identify with our data (Bielby and Baron, 1984). Self-employed men in sales occupations with higher proportions of women are likely to be proprietors, while women are their employees. Clearly, researchers must use the most refined categories possible to reveal the true amount of segregation within occupations. Most notably, we found that worker characteristics are important predictors of the percentage of women in sales occupations. After controlling for occupational characteristics in 1980 and in 1990, women tend to be concentrated in occupations with male workers that employers typically consider less desirable-those who are nonWhite, unmarried, and younger. Moreover, sales occupations showing the greatest increase in women over time are those with higher proportions of these workers. This finding is consistent with segmentation and queuing perspectives which suggest that employers rank and segregate workers by race, marital status, and age, as well as by sex. While one might argue that age is a proxy for experience and therefore represents a relevant criterion on which to sort workers, race and marital status are less easily dismissed. Moreover, these worker characteristics influence occupational sex composition after controlling for other characteristics of occupations and incumbents. Before labor analysts disregard Milkman's criticism that segmentation theorists do not differentiate between women and other sources of cheap labor, the possible intersection of gender, race, marital status, and age must

94

Bellas and Coventry

be examined for occupations other than sales. Based on this study's findings, we conclude that a multidimensional queuing theory may provide the best basis for such research. Reskin and Roos (1990; also see Sokoloff, 1992) suggest that employers rank workers on both race and gender, with White men tending to be most preferred, women of color rated least desirable, and White women and men of color ranked in the middle. While occupational racial segregation has received some attention in the literature (e.g., Tomaskovic-Devey, 1993; Sokoloff, 1992), researchers have paid little attention to segregation based on age or marital status (see Hutchens, 1988, for an exception). Since different occupations draw employees from different labor markets, future research should explore the extent to which the intersection of gender, race, age, and marital status occur in other types of labor markets. Perhaps employers in other occupations treat workers differently based on only one or two of these worker characteristics, rather than all of them. In addition to including age and marital status in subsequent analyses, researchers might also explore measures of contingency work. The extent to which full-time, year-round work is available is another factor that might influence occupational sex composition and should be examined in future research. Finally,we examined sales occupations in the expanding service sectorexpansion that was aided by women's increased labor force participation. We found that during the 1980s, sales occupations experienced a dramatic decline of over 11 points in the index of segregation. Despite this decline, the variables that affected the percentage of women in sales occupations remained quite stable between 1980 and 1990. In light of these findings and the structural changes in the economy, future research should examine the extent to which these factors also influence the sex composition in other service-sector occupations and manufacturing occupations, and if variables, such as occupational growth, which did not significantly affect the percentage of women in sales occupations, are significant predictors for other occupations. These results could help us better understand sex segregation in occupations (and, conversely, integration), by identifying whether factors that significantly predict occupational sex composition vary between occupational groupings and/or economic sectors. REFERENCES
Amott, Teresa 1995 "Shortchanged:restructuringwomen's work." In Margaret L. Andersen and Patricia Hill Collins (eds.), Race, Class, and Gender: An Anthology: 207-217. Belmont, CA: Wadsworth. Antilla, Susan 1995 "Young White men only, please: Women accuse a broker of blatant

Determinants of the Sex Composition of Sales Occupations discrimination." The New York Times (April 26): C1, C3. Arrow, Kenneth 1972 "Models of job discrimination." In Anthony H. Pascal (ed.), Racial Discrimination in Economic Life: 83102. Lexington, MA: Heath. Bellas, Marcia L. 1992 "The effects of marital status and wives' employment status on the salaries of faculty men: the (house)wife bonus." Gender & Society 6:609-622. Benson, Susan Porter 1984 "Women in retail sales work: The continuing dilemma of service." In Karen Brodkin Sacks and Dorothy Remy (eds.), My Troubles are Going to Have Trouble with Me: Everyday Trials and Triumphs of Women Workers: 113123. New Brunswick, NJ: Rutgers. Berheide, Catherine White 1988 "Women in sales and service occupations." In Ann Stromberg and Shirley Harkass (eds.), Women Working: Theories and Facts in Perspective: 241257. Mountain View, CA: Mayfield Publishing. Bielby, Denise D., and William T. Bielby 1988 "She works hard for the money: Household responsibilities and the allocation of work effort." American Journal of Sociology 93:1031-1059. Bielby, William T., and James N. Baron 1984 "A woman's place is with other women: Sex segregation within organizations." In Barbara F Reskin (ed.), Sex Segregation in the Workplace: Trends, Explanations, Remedies: 2755. Washington, DC: National Research Council. Biggart, Nicole Woolsey 1996 "Family, gender, and business in direct selling organizations."In Cameron Lynn Macdonald and Carmen Sirianni (eds.), Working in the Service Society: 157-183. Philadelphia, PA: Temple University Press. Blau, Francine D. 1984 "Occupational segregation and labor market discrimination." In Barbara E Reskin (ed.), Sex Segregation in the Workplace: Trends, Explanation, Remedies: 117-143. Washington, DC: National Academy Press.

95

Blau, Francine D. and Marianne A. Ferber 1992 The Economics of Women, Men, and Work. Englewood Cliffs, NJ: PrenticeHall. Blau, Francine D., and Carol L. Jusenius 1976 "Economists' approaches to sex segregation in the labor market: An appraisal." Signs: Journal of Women in Culture and Society 1:181200. Bradley, Harriet 1989 Men's Work, Women's Work. Minneapolis: University of Minnesota Press. 1993 "Across the great divide: The entry of men into 'women's jobs'." In Christine L. Williams (ed.), Doing "Women's Work": Men in Nontraditional Occupations: 10-27. Newbury Park, CA: Sage Publications. Catalyst 1995 Knowing the Territory: Women in Sales. New York: Catalyst. Cockburn, Cynthia 1988 "The gendering of jobs: Workplace relations and the reproduction of sex segregation." In Sylvia Walby (ed.), Gender Segregation at Work: 29-42. Milton Keynes, UK: Open University Press. Cohn, Samuel 1985 The Process of Occupational Sex Typing. Philadelphia: Temple University Press. Corcoran, Mary, Greg Duncan, and Michael Ponza 1984 "Work experience, job segregation, and wages." In Barbara F Reskin (ed.), Sex Segregation in the Workplace: Trends, Explanation, Remedies: 171-191. Washington, DC: National Academy Press. Cotter, David A., Joann M. DeFiorre, Joan M. Hermsen, Brenda Marsteller Kowalewski, and Reeve Vanneman 1995 "Occupational gender desegregation in the 1980s." Work and Occupations 22:3-21. Dassbach, Carl H.A. 1986 "Industrial robots in the American automobile industry."Insurgent Sociologist 13:53-61. Duncan, Otis D., and Beverly Duncan 1955 "A methodological analysis of segregation indices." American Sociological Review 20: 200-217.

96 Edwards, Richard 1979 Contested Terrain: The Transformation of the Workplace in the Twentieth Century. New York: Basic Books. Eitzen, D. Stanley, and Maxine Baca Zinn 1989 The Reshaping of America. Englewood Cliffs, NJ: Prentice Hall. England, Paula 1992 Comparable Worth: Theories and Evidence. New York: Aldine de Gruyter. Filer, Randall K. 1988 Data provided to Barbara A. Reskin and Patricia A. Roos for work on Job Queues, Gender Queues. Finch, Janet 1983 Married to the Job: Wives' Incorporation in Men's Work. London: Allen & Ulwin. Fowlkes, Martha R. 1980 Behind Every Successful Man: Wives of Medicine and Academe. New York: Columbia University Press. Game, Ann, and Rosemary Pringle 1983 Gender at Work. London: Pluto Press. Glass, Jennifer 1990 "The impact of occupational segregation on working conditions." Social Forces 68:779-796. Gordon, David M. 1972 Theories of Poverty and Underemployment. Lexington, MA: Heath. Hannan, Michael, and Alice A. Young 1977 "Estimation in panel models: results on pooling cross-sections and time series." In David R. Heise (ed.), Sociological Methodology: 52-83. San Francisco: Jossey-Bass. Hanushek, Eric A., and John E. Jackson 1977 Statistical Methods for Social Scientists. Orlando, FL: Academic Press. Hartmann, Heidi I. 1976 "The family as the locus of gender, class, and political struggle: The example of housework." Signs: Journal of Women in Culture and Society 6:366394. 1981 Women and Revolution: The Unhappy Marriage of Marxism and Feminism. London: Pluto Press. Hartmann, Heidi I., Kraut, Robert E., and Louise A. Tilly 1987 Computer Chips and Paper Clips: Technology and Women's Employment. Washington, DC: National Academy Press.

Bellas and Coventry Hutchens, Robert M. 1988 "Do job opportunities decline with age?" Industrial and Labor Relations Review 42:89-99. Jacobs, Jerry A. 1985 "Sex segregation in American higher education." In Laurie Larwood, Ann J. Stromberg, and Barbara A. Gutek (eds.), Women and Work: An Annual Review, vol. 1: 191-214. Beverly Hill, CA: Sage. 1989 Revolving Doors. Palo Alto, CA: Stanford University Press. 1992 "Women's entry into management: Trends in earnings, authority, and values among salaried managers." Administrative Science Quarterly 37:282301 Jacbbs, Jerry A., and Ronnie S. Steinberg 1990 "Compensating differentials and the male-female wage gap: Evidence from the New York State comparable worth study." Social Forces 69:439468. Kanter, Rosabeth Moss, and Barry A. Stein 1979 "The gender pioneers: Women in an industrial sales force." In Rosabeth Moss Kanter and Barry A. Stein Life (eds.), Organizations: Workplaces as People Experience Them: 134-160. New York: Basic Books. Karene, Ethel B., and Kenneth L. Anderson 1967 Women in Our Life-Insurance. Indianapolis, IN: Research and Review Service of America. Kokkelenberg, Edward C., and Donna R. Sockell 1985 "Union membership in the United States, 1973-1981." Industrial and Labor Relations Review 38: 497-543. Kuttner, Robert 1989 "The changing occupational structure." In D. Stanley Eitzen and Maxine Baca Zinn (eds.), The Reshaping of America: 85-102. Englewood Cliffs, NJ: Prentice Hall. Lawson, Helene M. 1993 "Car saleswomen: Expanding the scope of 'salesmanship'." Paper presented at the annual American Sociological Association meetings, Miami, Florida. Lieberson, Stanley 1980 A Piece of the Pie. Berkeley: University of California Press.

Determinants of the Sex Composition of Sales Occupations Loscocco, Karyn, Joyce Robinson, Richard H. Hall, and John K. Allen 1991 "Gender and small business success: an inquiry into women's relative disadvantage." Social Forces 70:6585. Macdonald, Cameron Lynn, and Carmen Sirianni 1996 "The service society and the changing experience of work." In Cameron Lynn Macdonald and Carmen Sirianni (eds.), Working in the Service Society: 1-26. Philadelphia, PA: Temple University Press. Marsden, Peter V., Arne L. Kalleberg, and Cynthia R. Cook 1993 "Gender differences in organizational commitment: Influences of work positions and family roles." Work and Occupations 20:368-390. Milkman, Ruth 1987 Gender at Work: The Dynamics of Job Segregation by Sex during World War II. Urbana, IL: University of Illinois Press. Mitchell, Juliet 1971 Woman's Estate. New York: Pantheon Books. 1974 Psychoanalysis and Feminism. New York: Pantheon Books. Polachek, Solomon 1979 "Occupational segregation among women: Theory, evidence, and a prognosis." In Cynthia Lloyd (ed.), Women in the Labor Market: 137-157. New York: Columbia University Press. Ransom, Michael R. 1990 "Gender segregation by field in higher education." Research in Higher Education 31:477-491. Reskin, Barbara E 1988 "Bringing the men back in: Sex differentiation and the devaluation of women's work." Gender & Society 2: 58-81. Reskin, Barbara F, and Irene Padavic 1994 Women and Men at Work. Thousand Oaks, CA: Pine Forge Press. Reskin, Barbara E, and Patricia A. Roos 1990 Job Queues, Gender Queues. Philadelphia: Temple University Press. Sokoloff, Natalie J. 1980 Between Money and Love: The Dialectics of Women's Home and Market Work. New York: Praeger.

97

1992 Black Women and White Women in the Professions. New York: Routledge. Steinberg, Ronnie J. 1990 "Social construction of skill: Gender, power, and comparable worth." Work and Occupations 17:449-482. Strober, Myra H. 1984 "Towards a general theory of occupational sex segregation: The case of public school teaching." In Barbara F. Reskin (ed.), Sex Segregation in the Workplace: Trends, Explanation, Remedies: 144-156. Washington, DC: National Academy Press. Strober, Myra H., and Carolyn L. Arnold 1987 "The dynamics of occupational segregation among bank tellers. " In Clair Brown and Joseph A. Pechman (eds.), Gender in the Workplace: 107-148. Washington, DC: Brookings Institution. Thomas, Barbara J. 1990 "Women's gains in insurance sales: Increased supply, uncertain demand." In Barbara F. Reskin and Patricia A. Roos (eds.), Job Queues, Gender Queues: Explaining Women's Inroads into Male Occupations: 183204. Philadelphia: Temple University Press. Thomas, Barbara J., and Barbara E Reskin 1990 "A woman's place is selling homes: Occupational change and the feminization of real estate sales." In Barbara F. Reskin and Patricia A. Roos (eds.), Job Queues, Gender Queues: Explaining Women's Inroads into Male Occupations: 205-223. Philadelphia: Temple University Press. Thurow, Lester 1969 Poverty and Discrimination. Washington, DC: Brookings Institution. 1975 Generating Inequality. New York: Basic Books. Tomaskovic-Devey, Donald 1993 Gender and Racial Inequality at Work: The Sources and Consequences of Job Segregation. Ithaca, NY: ILR Press. Treiman, Donald J., and Kermit Terrell 1975 "Women, work, and wages-trends in the female occupation structure." In Kenneth C. Land and Seymour Spilerman (eds.), Social Indicator Models: 157-199. New York: Russell Sage Foundation.

98 Urquhart, Michael 1984 "The employment shift to services: where did it come from? Monthly Labor Review 107:15-22. U.S. Bureau of the Census 1983 Census of Population and Housing, 1980: Public-Use Microdata Sample. Washington, DC: The Bureau of the Census. 1993 Census of Population and Housing, 1990: Public-Use Microdata Sample. Washington, DC: The Bureau of the Census.

Bellas and Coventry 1996 Statistical Abstract of the United States: 1996. Washington, DC: U.S. Government Printing Office. U.S. Department of Labor, Bureau of Labor Statistic 1996 Occupational Outlook Handbook, 1995-96 edition. Washington, DC: U.S. Government Printing Office. Williams, Christine L. 1989 Gender Differences at Work: Women and Men in Nontraditional Occupations. Berkeley: University of California Press.

S-ar putea să vă placă și