HUMUHWNHIWHIWMll‘lelNWll n 1 WI N_\ #N #N '—4 I 'm “we as j ' l Pfiié: 7‘ 1.. ;.j". 1:. v University L J- ——7 This is to certify that the thesis entitled WOMEN'S OCCUPATIONS WITHIN A DUAL LABOR MARKET FRAMEWORK: A STUDY OF SINGLE AND MARRIED WOMEN'S OCCUPATIONAL MOBILITY presented by VIRGINIA E . POWELL has been accepted towards fulfillment of the requirements for MASTER— degree in _AR:ES__ SOCIOLOGY CW 7C 740%, Jajor professor Date_D_eg.emheL_22._1_9 8 0 0-7639 'I/lllllllllllllll V‘Q. OVERDUE FINE: . a A. W per mm C ’ .‘TERlflLii G LIBRARY W ._... (KI/(my . SELL—~— -.4' . um 30 ’9"3".’3 Place 1" bOOk-?t1dr';0n regc, ‘15 . ”.2." me from CHCU' "’ ‘chi‘ w “,,I” (VJ! Cha 3 v 4:- ~ ”I _ ' 'FQ 8243:“; q WOMEN'S OCCUPATIONS WITHIN A DUAL LABOR MARKET FRAMEWORK: A STUDY OF SINGLE AND MARRIED WOMEN'S OCCUPATIONAL MOBILITY BY Virginia E. Powell A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF ARTS Department of Sociology 1981 ABSTRACT WOMEN'S OCCUPATIONS WITHIN A DUAL LABOR MARKET FRAMEWORK: A STUDY OF SINGLE AND MARRIED WOMEN'S OCCUPATIONAL MOBILITY BY Virginia E. Powell This study uses a dual labor market approach to examine married versus single women's intragenerational occupational mobility. Primary jobs are characterized by upward mobility, while secondary jobs are characterized by no mobility or downward mobility. It is hypothesized that single women are more likely than married women to be employed initially in primary occupations, and are thus upwardly mobile. Married women, it is hypothesized, will be employed initially in secondary occupations, and are thus not mobile or are downwardly mobile. Using contingency tables to analyze mobility for a sample of 1794 women, it was found that married women were more likely to be initially employed in a primary occupation, and thus potentially upwardly mobile. Single women, however, had higher rates of actual mobility over time when compared with married women. Two conclusions Virginia E. Powell were drawn: a dual labor market framework can only partially account for the experience of women in the labor force; and, marital status is an important variable in any consideration of occupational mobility. ACKNOWLEDGMENTS This completed project is one not done in isolation, and there are many people to whom sincere thanks are offered. My advisor, Craig Harris, was very supportive and encouraging of my work. His speedy reading and commenting on rough drafts was greatly appreciated. Thanks also to my committee members, Nan Johnson and Jim Zuiches, for their careful reading and recommendations on the thesis. Many thanks go to Jody Crawford and Paul Winder, who helped me with the computer programming. Mary McCormack was a dear friend and a loving critic during the early stages of my research. Much thanks to her for her comments on writing style and her very warm support. Finally, I express sincere appreciation to Dale Jager for his assistance in computer programming and, more importantly, his moral support--a kind of intellectual and personal Vitamin C. ii CHAPTER I: CHAPTER II: TABLE OF CONTENTS Page INTRODUCTION.0.0.0.0000....0... 1 WOMEN IN OCCUPATIONS: WHAT PRICE MOBILITY?........... 4 Patterns of Female Employment........... 4 The Dual Labor Market Perspective of Economic Discrimination.............. 12 Women's Occupational Mobility........... 25 CHAPTER III: CHAPTER IV: CHAPTER V: APPENDIX: BIBLIOGRAPHY: RESEARCH DESIGN AND DATA DESCRIPTIONS.COOOOOOOOOOOO 37 DATA ANALYSISOOOOO0.00.00.00.00 43 CONCLUSION AND DISCUSSION...... 63 0.00.0...OOOOOOOOOOOOOOOCOOOOOO 68 00.0.0...OOOOOCOOOOOOOOOOOOOOOO 76 iii Table Table Table Table Table Table Table Table Table Table 10. LIST OF TABLES Distribution by Sex in Census Occupational Categories, 1977 Median Years on Current Job by Age and Sex, 1973 Women Excluded from Sample for Current Analysis First Occupation Mobility Potential by Marital Status Actual Mobility from First Occupation to 1974 Occupation by Marital Status First Occupation Mobility Potential by Actual Mobility from First Occupation to 1974 Occupation for Married Women First Occupation Mobility Potential by Actual Mobility from First Occupation to 1974 Occupation for Never Married Women Actual Mobility from First Occupation to 1974 Occupation by Marital Status, for those Women Whose First Occupation Mobility Potential is "High Move Up" Actual Mobility from First Occupation to 1974 Occupation by Marital Status, for those Women Whose First Occupation Mobility Potential is "Low Move Up" Actual Mobility from First Occupation to 1974 Occupation by Marital Status, for those Women Whose First Occupation Mobility Potential is "No Move" iv Page 41 45 47 49 50 53 54 55 Table Table Table Table Table Table Table Table Table 11. 12. 13. 14. 15. 16. 17. 18. 19. Actual Mobility from First Occupation to 1974 Occupation by Marital Status, for those Women Whose First Occupation Mobility Potential is ”Low Move Down" Actual Mobility from First Occupation to 1974 Occupation by Marital Status, for those Women Whose First Occupation Mobility Potential is "High Move Down" Actual Mobility from First Occupation to 1974 Occupation by Marital Status, for those Women Whose First Occupation Mobility Potential is "Primary" Actual Mobility from First Occupation to 1974 Occupation by Marital Status, for those Women Whose First Occupation Mobility Potential is "Secondary" Census Occupations with a High Degree of Upward Mobility Census Occupations with a Low Degree of Upward Mobility Census Occupations with No Mobility Census Occupations with a Low Degree of Downward Mobility Census Occupations with a High Degree of Downward Mobility 56 57 59 60 68 70 72 74 75 CHAPTER I INTRODUCTION In February of 1960, four black men sat down at a Woolworth's lunch counter in North Carolina and ordered coffee. They were refused service on account of their color. This refusal became "the shot heard round the world" of the Civil Rights Movement, and triggered the subsequent black insurgency in American society. Concern with the issue of race and racism spread to the academic world as social scientists began to question the sources of racial inequality. Two possible sources which came under close scrutiny were family and work. Perhaps a pathology or deterioration within the family would explain the position of blacks in society, or perhaps the problem of racial inequality was due to structural inequality, particularly in institutions like the economy and education. In August of 1970, tens of thousands of women marched down Fifth Avenue in New York City, and created in their wake the second feminist movement. With their fists and their consciousness raised, these women sought to awaken people to the sources of sexual inequality in American society. And just as the Civil Rights Movement forced 1 academicians to reexamine the bases of racial inequality, the women's movement forced academicians to investigate commonalities in the origins of racial and gender inequality. Consequently, social scientists explored the family and work as key sources of gender inequality. As a sociologist coming of age during the 19705, I became aware of both the strengths and limitations of the literature being written about women. Of particular interest to me was the place of women in the world of work. For it seemed that during the past decade, a contradition-- which had been present in American society--came to the center. With continued inflation, it became necessary in many households for both husband and wife to work. Where prior ideological definitions had restricted many women to the world of family concerns and excluded them from the world of competitive employment, material needs were forcing a large percentage of them into the work force. And yet, upon entering the world of work, women were experiencing a very different world from that of men. Women systematically held lower status, lower paying jobs. In order to explain this phenomenon, sociologists examined occupational structures and rates of mobility across these structures. But the work of these scholars has ignored a fundamental and essential question: How does marital status influence the placement of women in occupational categories and their subsequent mobility? Because I believe that how questions are asked, and the way in which they are framed, and the manner in which they elicit information is essential, I turned my attention to this question. In this way, I see myself as following in the footsteps of those who have blazed a path into the subject of sexual inequality, and also as taking one step back to seek an understanding of a fundamental issue. CHAPTER II WOMEN IN OCCUPATIONS: WHAT PRICE MOBILITY? Patterns of Female Employment (Since World War II, the extent of women's participation in the labor force has changed drastically. The socio- demographic characteristics of women workers have also shifted during this time. It is therefore important to examine the most salient trends in female employment in order to assess their impact on the current labor market structure. The single most important feature to emerge about women's work iftes in the 19605 and 19705 is its sheer X! ~t, volume. In 1977, 48(percent of all women aged 16 and over fr /" .,/ were in the labor force. In turn, women accounted for 41.1 percent of the total labor force (United States Department of Labor, 1977,-p. A-l9, 52)./)Seventy-one percent of women who worked were full-time employees. The largest increase in labor force participation rates between 1960 and 1977 was among those women aged 20 to 34, while from 1950 to 1959 the largest increase had been among women aged 45 to 64. These figures show that, besides the fact that women are entering the labor force in higher numbers, they are also entering at younger ages (Howe, 1977, Appendix--Chart C and F). Most women who work are married. In 1974, married women accounted for 56.8 percent of all working women. Of all married women, 46.6 percent were in the labor force. Married women with children are more likely to work today than ever before. In 1956, 20 percent of all women with children under six years of age were working; in 1967, that figure was 29 percent; and in 1977 the rate was 40.9 percent. In turn, 40 percent of all women with children from 6 to 17 years of age were in the labor force in 1956; in 1967, their participation had risen to 49 percent, and had increased still more to 58.3 in 1977 (Vogel, n.d., pp. 6-7; Howe, 1977, Appendix--Chart E; United States Department of Labor, 1977, pp. 52-53). Ilene Winkler has ventured an explanation for the increased participation of women in the labor force, particularly those who are married with children, in the past decade. She suggests that historically women were used in times of war and labor shortages as a reserve army, and moved in and out of the labor force to accommodate these fluctuations. More recently, however, women have entered the world of work to supplement the family income, and to alleviate some of the burdens "because of recessions in the 19505 and inflation in the 19605, because of rising taxes and the squeeze caused by credit buying." Winkler points out that in 1965, seventy-five percent of all married women workers came from families in which their husbands were earning less than $7000.00 annually, at a time when the federal government was citing that sum as a modest budget for a family of four. Approximately 30 percent of all married men earned less than $7000.00 annually (Winkler, n.d., pp. 2-3)o Table l: DISTRIBUTION BY SEX IN CENSUS OCCUPATIONAL CATEGORIES, 1977 Census Category Men Women Description Professional, technical, and kindred 13.0% 13.5% Farmers and farm managers 1.8% 0.4% Managers, officials, and proprietors, except farm 11.8% 6.5% Clerical and kindred workers 6.5% 30.6% Sales workers 5.7% 6.0% Craftsmen, foremen, and kindred 20.6% 2.1% Operatives and kindred 19.7% 13.8% Private household workers 0.1% 4.6% Service workers 11.1% 21.2% Farm laborers and foremen 2.3% 0.3% Laborers, except farm and mine 7.5% 1.1% Source: United States Department of Labor, Bureau of Labor Statistics-- March, 1977 When women do go to work, they earn substantially less than their male counterparts. Nancy Barrett showed that in 1975, women earned 58.8 percent of what men earned. The largest employer of women, the clerical field, paid women only 62.2 percent of the wages paid to men in the same field. Further, this discrepancy between women's and men's earnings is not accounted for by differences in job assignment (Barrett, in Smith, 1979, p. 34). Women and men do not generally hold the same kind of jobs. The labor market structure is decidedly split along sex lines. In 1977, 79.1 percent of all employed women were found in four of the eleven Census occupational categories. Table 1 shows the distribution of women and men in these major categories. The index of dissimilarity between these two distributions is 39.5, which indicates that there is a substantial difference in allocation by sex to these categories. Women dominate in the "clerical and kindred" and "service workers" categories. Men are more likely found in the "craftsmen, foremen, and kindred workers" category. These two disparities account for seven-eighths of the dissimilarity indicated by the index. As we will see later in this chapter, this distribution may be slightly deceptive because it does not show intra-categorical differences in occupations by sex. For example, women and men are represented almost equally in the "professional, technical, and kindred workers" category. Within this category, however, women are more likely than men to be elementary school teachers and registered nurses, while men are more likely to be college professors and doctors. Using such broad categories, then, may mask job segregation by sex, which would be more apparent if categories were more narrowly defined. Valerie Oppenheimer has shown that the occupations women usually hold are those for which skill is required prior to employment. For example, teachers and nurses must obtain education beyond the high school level in order to learn their trades. Secretaries, stenographers, and typists must acquire their specific skills before seeking employment. In short, the jobs that women tend to fill offer little or no on-the-job training, instead requiring that skills be learned before employment is granted (Oppenheimer, 1970). Ilene Winkler has said that this is a very profitable way for industries and employers to handle the work women do. A woman's work is generally not continuous throughout her life. She may drop in and out of the labor force to get married, to move with her husband to his new job, to bear and raise children, or to accommodate the financial needs of the family. This pattern has been documented by the United States Department of Labor. In 1973, the median number of years spent at a job was 2.8 years for women and 4.6 years for men (Winkler, n.d.). Further, this discrepancy is sometimes greater when age cohort is controlled. Table 2 illustrates the pattern by age. Since it is true that women do not stay as long at a job, it is not in the interest of an employer to "invest" a great deal of training in a person whose tenure at any given job is questionable. A high rate of turnover in some occupations would represent a high cost to that employer. Women are therefore restricted to those jobs which require low investment, in terms of on-the-job training, to employers (United States Department of Labor, 1976). Table 2: MEDIAN YEARS ON CURRENT JOB BY AGE AND SEX, 1973 Total, 16 years and older 2.8 4.6 16-24 years 0.9 . 0.8 25-34 years 2.2 3.2 35-44 years 3.6 6.7 45-54 years 5.9 11.5 55-64 years 8.8 14.5 65 years and over 10.9 13.9 Source: United States Department of Labor, Women's Bureau-- 1976. This intermittent employment pattern for women can also explain why women are not granted jobs with seniority rights. Since it is believed by employers that women are not stable employees, those employers in need of relatively 10 permanent employees to reduce training costs are less likely to grant seniority rights to women than to men. They are also overlooked more frequently for promotions because it is believed that their continuance at a place of work is tenuous. A woman with children is often passed over due to her employer's assumption that she will have a high rate of absenteeism because of her children. If a woman is single, she is considered a bad risk in that she might get married and have to move with her husband (Winkler, n.d.; Davies, 1974). This ideology relegates women to certain occupations, those with low employer investment, while restricting them from others. This relegation of women to certain occupations is not due to women's lack of education. In 1974, both men and women had the same median number of years of schooling-- 12.3. Employed women and men had a median of 12.5 years of school completed. For persons not in the labor force, women had completed more years of school than men--12.0 years and 10.3 years, respectively. While these figures do not control for content of education, they nonetheless point out that women have at least as much education as men. Women do not, therefore, bring less generalized human capital to the job market. Women are not only discriminated against occupationally. The jobs they do hold are less likely to be represented by labor organizations, and women are less likely to become members of unions. Women constitute 41.1 11 percent of the labor force, but only 27.6 percent of persons represented by labor organizations. In turn, 15.7 percent of all working women are represented by labor organizations, compared to 23.8 percent of men in the labor force. Workers represented by labor organizations earn more than those who do not. Women's low participation rates in labor unions, then, would contribute to an explanation of the earnings gap between men and women (United States Department of Labor, 1979). In summary, seven important trends have been noted in female employment in the United States: 1) women are becoming a larger proportion of the labor force, particularly married women with children under the age of 17; 2) women earn substantially less than men in all major occupational categories; 3) women and men do not work in the same jobs; 4) women's work is concentrated in those jobs which require little or no on-the-job training costs to employers; 5) women spend less time working in one job than men; 6) women and men have the same median level of education; and, 7) women are less likely than men to be members of labor organizations. The dual labor market perspective is one framework which can be used to describe the interactive operation of these trends. The Dual Labor Market Perspective of Economic Discrimination Dual labor market theory was developed in the 19605 in an attempt to explain why the Federal government's "war on poverty" had failed. The human capital school of economics had proposed that the poor were poor because they lacked marketable skills with which they could sell themselves in the labor force. The differences in income between the poor and the middle class reflected the differences in education, training, work experience, health, and subsequent mobility between the two groups. These ideas in turn led to the institution of manpower programs which attempted to rehabilitate the working poor by providing on-the-job training and skills. These programs failed to alleviate the burden of poverty, and were criticized by some economists, psychologists, and sociologists because they had failed to see that the problem was not one of the individual worker, but rather a problem in the American economy. The human capital school had been treating symptoms, not causes, in their proposed solutions (Bluestone, 1970). One of the theoretical frameworks to emerge from these critiques was the dual labor market theory, whose main proponents are Peter B. Doeringer and Michael J. Piore. In their book Internal Labor Markets and Manpower Analysis, Doeringer and Piore define two distinct types of 12 13 jobs: primary sector jobs and secondary sector jobs. Primary sector jobs are characterized by high wages, good working conditions and employee benefits, employment stability, chances for advancement, equity and due process in the administration of work rules, and a high rate of union membership. Secondary sector jobs, by distinction, generally have the following characteristics: low wages, little or no fringe benefits, poor working conditions, high labor turnover, little chance for advancement, arbitrary and random administration of work rules, and little or no union membership. Workers are generally confined to the secondary sector by residence, inadequate skills, poor work histories, and most importantly, discrimination (Doeringer and Piore, 1971). The most critical distinction to be made between the primary and secondary sector is the development in the primary sector of internal labor markets and the lack thereof in the secondary sector. Michael L. Wachter defines an internal labor market as: "...a set of structured employment relationships within a firm, embodying a set of rules, formal or informal, that govern each job and their interrelationships. These rules...cover job content and wages, opportunities on the promotion ladder, and grievance procedures. A complex employment relationship has developed primarily because of the elaboration of tasks that are specific to a job and hence require specific training, often acquired on the job. New workers are used principally to fill entry jobs, while most higher-level 14 positions are filled by promotion from within" (Wachter, 1974, pp. 642-643). Within the secondary sector, external labor markets exist. An external labor market does not have structured relationships or rules governing each job level. There are no tall promotion ladders attached to job categories: employment in the sector is responsive to forces of supply and demand in the external market. There are not many specific skills required for any job, and workers are easily replaceable in a job (Doeringer and Piore, 1971; Wachter, 1974). The internal/external labor market structure overlaps both primary and secondary job categories. That is, internal labor markets exist in secondary sector jobs and external labor markets exist in primary sector jobs. But the point is that, in most cases, primary jobs are part of an internal structure because they require skill specificity, and thus the employer benefits from low job turnover. Secondary jobs have external structures because their need for skills and thus job continuity are minimal. Corresponding to the primary/secondary distinction in job categories is a distinction between primary and secondary industries. Robert T. Averitt makes this distinction in his book The Dynamics of American Industry Structure. According to Averitt, primary sector industries or firms are characterized by: 1) a large economic size, as measured by number of employees, total assets, and yearly 15 sales; 2) geographic dispersion, both nationally and internationally; 3) product diversification; 4) managerial decentralization; and, 5) concentrated markets. By contrast, secondary sector industries or firms have the following characteristics: 1) a small economic size, with limited growth potential; 2) little or no geographic dispersion, either nationally or internationally; 3) production of only a small line of related products; 4) centralized management, often around one or two individuals; and, 5) unconcentrated markets (Averitt, 1968). Averitt suggests that the life span of each industry is determined by these characteristics. Primary sector industries are economically stable, and so can plan for growth over a long period of time. Their high profits allow them to hire workers, and offer them higher wages and good benefits, while also promising some opportunity for movement. Secondary sector industries, however, are not stable economically, and so offer very little to potential employees in terms of wages, benefits, and mobility (Averitt, 1968). Therefore, both industries and occupations can be distinguished as either primary or secondary in nature. Doeringer and Piore cite three distinct factors that contribute to the emergence and maintenance of internal labor markets. Primarily, the need for Specific skills in any given occupation increases training costs for the employer. It is thus in the economic interest of the 16 employer to discourage job turnover when s/he "invests" in a worker and provides job specific skill training. The employer will attempt to minimize turnover by providing long term benefit promises to the worker in the form of seniority and promotion rights, higher wages if one advances from one job to another within the firm, and other highly desirable fringe benefits. Secondly, on-the-job training provides the worker with skills actually utilized in a particular job that cannot be learned by the employee in school. Piore and Doeringer explain that this second factor is difficult to define because the worker learns many skills just by observing someone else performing them, seemingly by a process of "osmosis." The importance of this osmosis is that those skills learned by observation may later earn the worker a promotion because s/he is the natural choice to fill a particular position. S/he is the most readily available person who knows that skill. And for most of these types of jobs, there is no alternative to on-the-job training. Thus, an informal contract is made between worker and employer. The worker learns job specific skills which are not readily transferable outside of the firm, and which will later be used as criteria for seniority and promotion within that firm. S/he, in short, accepts the promise of long term benefits sometime in the future. The employer, in turn, has purchased some commitment for long term employment by the worker which will reduce her/his turnover costs, at the expense of offering a higher quality package of benefits l7 (Doeringer and Piore, 1971). Finally, Doeringer and Piore discuss the role of custom in maintaining this contract. Custom behaves like a stabilizer in the workplace in its setting up of rules and norms for behavior--ones which apply to both employer and worker. Rules governing seniority and promotion, as Doeringer and Piore point out, are particularly sensitive to custom (Doeringer and Piore, 1971). The most important feature of primary occupations, and thus internal labor markets, is that they tend to develop mobility clusters. Mobility clusters are defined lines of job progression. Job skills fall into a natural skill progression, where the training required to learn any given job in the progression occurs in the job below it. A worker accumulates skills and moves up in the progression, drawing upon her/his seniority. Jobs that fall into a progression, and there are many of these within any given firm, form a mobility cluster. Theoretically, a worker who faithfully learns skills in the job progression is essentially guaranteed by custom a move within the cluster. This insulates the worker from any competition external to the firm (Doeringer and Piore, 1971). Harrison White, in his book Chains of Opportunity, proposes a model of mobility which can explain the development of mobility clusters. White says that there are two dimensions to any mobility system: a set of fixed, interrelated jobs, and a set of people qualified for these 18 jobs. Jobs within this system are stratified by skill requirements, and people are stratified by skill qualifications. Mobility chains, or vacancy chains, are established when a person leaves the system and creates a job vacancy. Another person in the system must fill this vacancy, and is chosen by two criteria--promotability, in terms of skill qualification; and seniority. The vacancy chain has begun, and each job will be filled from below until someone new enters the system at the bottom (White, 1970). White suggests that this model can be used with probability theory to determine the structure of opportunity at two levels: primarily, it can illustrate mobility potentials within a set of job categories, to learn the structural potential for movement; and secondly, it can trace individual careers within a mobility system by measuring chances for movement (White, 1970). White's model could therefore be viewed as an elaboration on mobility clusters. His theoretical framework would aid in establishing, for any given firm, the networks of job categories and how they contribute to, or restrain, movement. According to Piore and Doeringer, entry into the primary labor market and/or internal labor market, and thus possible access to a job in a mobility cluster, is governed by specific criteria, the most important of which is education. Education is a discriminatory tool used by 19 employers to identify undesirable employees who have not attained a certain level of formal schooling. Less important criteria include aptitude test scores, personal interviews, work experience, physical fitness, and any other characteristics seen as desirable by management (Piore and Doeringer, 1970). Randall Collins has challenged the assumption that level of education defines occupational capabilities for all people. Using case studies and cross-sectional studies, Collins suggests that education in an advanced industrial society is a way for status groups to maintain their privilege. Education does not necessarily act to better prepare people for highly technological jobs, but rather is a way to sort into high status, high paying jobs those people who are, figuratively, ”protestant, male, and completely white," i.e., those people whose social characteristics mirror the business elite (Collins, 1971, p. 1008). Education, in short, marks people as members of a particular group, and is not a mark of technical skill and achievement. Collins summarizes the use of education in the United States in this way: "...the evidence indicates that educational requirements for employment reflect employers' concerns for acquiring respectable and well-socialized employees; their concern for the provision of technical skills through education enters to a lesser degree. The higher the normative control concerns of the employer, and the more elite the organization's status, the higher his 20 educational requirements" (Collins, 1971, p. 1014). Thus, education is a sorting mechanism used by employers to insure "good" workers whose background is similar to their own. The attainment of job-specific skills is a less important factor. Doeringer and Piore use dual labor market theory to analyze black unemployment and underemployment. They suggest that blacks are confined to the secondary sector job categories largely because of racial discrimination. Employers are able to choose workers for jobs along racial lines, and support their decisions by citing blacks as having low commitment to work, poor work histories, high rates of absenteeism and lateness, and inadequate skills. Another factor in that block, however, is the belief by employers that the general education level of blacks is lower than that for whites. Whether or not these differences are real, blacks end up in secondary sector jobs (Doeringer and Piore, 1971). These authors continue by describing the nature of jobs in the secondary sector, and their relationship to racial discrimination. Jobs in the secondary sector, they suggest, require little or no on-the-job training and offer limited possibilities for career advancement. These conditions reduce a worker's incentive to remain on the job and perform well. The wage paid is not likely to be higher than the wage in any other secondary sector job, and there is only a 21 slight chance that this wage would be increased through promotions or seniority. An employer, in turn, is less reluctant to lay off a worker in the secondary sector, since a large investment has not been made in training. So workers in this secondary sector, with secondary status jobs, exhibit higher rates of turnover, absenteeism, lateness, petty theft, and insubordination relative to workers in the primary sector. The relative lack of rewards for employment, it is argued, results in a very high rate of unemployment among minorities, specifically blacks (Doeringer and Piore, 1971). Doeringer and Piore contend that the positions of blacks in the labor force can be attributed more to sector and job segregation than to a lack of skills or education. They also point out that employers discriminate on the basis of perceptions of black workers, without considering how the structure of opportunities for blacks perpetuates these perceptions (Doeringer and Piore, 1971). The proposal of Doeringer and Piore that the inferior status of blacks in the labor market is more due to sectoral and job segregation than to individual differences between blacks and other workers has been empirically tested and verified. Elwood Beck, Patrick Horan, and Charles Tolbert studied the relationship between labor market segmentation and discrimination against non-whites and women. More specifically, they investigated the relationship between the structure of primary and secondary sectors, and the 22 perpetuation of the inferior status of minorities through discrimination. They attempted to measure two discriminatory mechanisms: the differential allocation of minorities to sectors, and the differential evaluation of worker credentials within each sector (Beck, et. al., 1978b). In order to test differential allocation by sector, Beck et. a1. divided sectors by three criteria: 1) economic scale of an industry (assets, number of workers, profit); 2) product market strength (product concentration); and, 3) labor market factors (percent unionization, percent full-time workers, hourly wage, weekly wage). Their analysis showed that women were more likely than men (both white and non-white) to be located in the secondary sector industries, and that non-white males were more likely than white males to be allocated to the secondary sector (Beck, et. al., 1978b). Beck, Horan, and Tolbert then tested to see if differential allocation by race and sex was due to direct discrimination by employers, or discrimination which resulted from minorities' acquisition of lower levels of human capital-~schooling and labor market experience. The authors concluded from their analysis that human capital deficits cannot completely explain sectoral segregation. Direct discrimination, based on sex and race, contributes more to an explanation of sector allocation of minorities than do either schooling or work experience (Beck, et. al., 23 1978b). Francine D. Blau and Carol L. Jusenius tried to determine the theoretical goodness of fit between the experience of women in the labor force and dual labor market theory. That is, they examined job segregation by sex to see if it could be explained by dual labor market theory. They point out that not only are occupations segregated by sex interoccupationally, but also intraoccupationally, within a dual labor market structure. Females occupy different jobs than males do, and these jobs are more likely to be of a secondary sector type, or those primary sector jobs which do not have a highly developed internal labor market. Blau and Jusenius contend that employers segregate jobs for women primarily because women do not usually work continuously like men do. They drop in and out of the labor force to accommodate family and life cycle demands. Therefore the jobs they do, according to employer interests, must be those which demand little skill specificity and on- the-job training costs to the firm. And, according to employer interests, these jobs pay low wages and offer few benefits, and include such jobs as clerk, salesperson, secretary, and typist, i.e., occupations dominated by females (Blau and Jusenius, 1976). Blau and Jusenius also suggest that within occupational categories, sex segregation is further extended so that, for example, women salespersons in a department store sell apparel and small appliances for hourly wages, while their 24 male counterparts sell furniture and large appliances for which they receive a commission on their salary (Blau and Jusenius, 1976). Blau had made an empirical test of intraoccupational segregation, and found that when she controlled for skills and abilities between men and women, a strong and consistent pattern of sex segregation within firms still existed (Blau, 1973). James W. Grimm and Robert N. Stern found that even in fields where women dominate as a numerical majority--those of nursing, school teaching, librarianship, and social work-~work was sex segregated, with men disproportionately in positions of supervision and planning (Grimm and Stern, 1974). Women can thus find employment in either primary or secondary sector occupations. When working in the primary sector, however, they acquire those jobs which have short career ladders attached to them, and resemble most closely secondary sector jobs. This has serious consequences for women's mobility within the labor force. The issue of how women get distributed into these types of jobs remains unanswered. It is, however, a problem of direct interest here, but one which cannot be dealt with until we have considered the literature on occupational mobility, since the structure of opportunities for mobility is intrinsic to dual labor market theory. It is to this task that we now turn. Women's Occupational Mobility Studies of occupational movement are either macro— structural or micro-structural. Macro-structural analyses attempt to outline the structure of occupational Opportunity within a social system. They examine the relationship between supply and demand for occupational categories, and the degree of movement in and out of these categories over time. One of their primary concerns is with factors affecting patterns of occupational mobility and individual chances for success. The classic example of a macro-structural examination of mobility is Peter M. Blau and Otis Dudley Duncan's The American Occupational Structure. Using a sample of over 25,000 American men between the ages of 20 and 64, Blau and Duncan measured occupational mobility in two ways. Primarily, they measured intergenerational mobility from respondent's father's occupation, to respondent's first occupation and occupation in 1962. This was done in order to discern patterns of Opportunity intergenerationally, and to observe changing trends in the occupational structure. Secondly, Blau and Duncan measured intragenerational movement from respondent's first job to 1962 occupation, again to discern opportunities for movement within a changing occupational structure (Blau and Duncan, 1967). A large part of their analysis was thus concerned with the 25 26 changing nature of the occupational structure on a macro- level, although they also used their data to describe the status attainment of individuals at a micro-structural level. David L. Featherman and Robert M. Hauser recently replicated Blau and Duncan's research. In 1973, they interviewed approximately 33,600 men to investigate further changes in the American occupational structure, and compared their findings with those of Blau and Duncan. Within certain constraints imposed by reinterviewing methods, Featherman and Hauser say their research is a definite replication, and proceed to analyze mobility intergenerationally and intragenerationally (Featherman and Hauser, 1978). Macro-structural measures of mobility focus on the changing nature of occupational opportunities both intergenerationally and intragenerationally. Such studies are concerned with the construction of matrices of employment opportunities over time, and seek patterns of occupational mobility over time. Blau and Duncan, and Featherman and Hauser's research ventures are two representatives of this type of research. Micro-structural measures of mobility are more concerned with the status attainment of individuals than with structural factors in movement. These mobility studies measure the impact of an individual's socio-demographic characteristics--social origins, education, marital status, 27 age, sex, race-—on her/his occupational achievement. Micro- structural studies are also done both intergenerationally and intragenerationally. Blau and Duncan's research was once again the pacesetter in this area of mobility studies. Intergenerational studies will not, however, be the main focus for this analysis. This study will utilize the dual labor market theory as a framework from which toevaluate the distribution of positions of a select group of women in primary versus secondary occupations. Further, the central question to be addressed is whether or not marital status has an impact on the distribution of women across different types of jobs; and, if so, what implications this distribution has for the occupational movement of single versus married women intragenerationally. There are two reasons why an intergenerational approach is inadequate here. Primarily, intergenerational mobility studies tend to mask job segregation by sex. Because they measure broad patterns of movement between two generations, their results sometimes camouflage the differentiation of the labor force that exists. This is nowhere more apparent than in the work of Peter Y. De Jong and his colleagues. De Jong et. al. used a sample of 2371 females aged 21 and over who had ever worked, and compared their intergenerational movement to males in Blau and Duncan's study. They predicted that, because of women's conflicts between domestic and career roles, and the extreme pressure placed on women to maintain 28 domestic obligations as primary, women would experience downward mobility occupationally when compared with their fathers. This hypothesis was not substantiated by their data. They found that the same general pattern exists for males and females with respect to mobility: upward mobility was more prevalent than downward mobility, short distance movement occurred more often than long distance moves, and barriers to downward movement were stronger than barriers to upward movement (De Jong, et. al., 1971). De Jong et. al. used ten broad census categories to define occupation. Their analysis masks the fact that, within each of these categories, men and women have different jobs, and that women typically have the lower status, lower paying jobs within each category. Thus, while their conclusions may be correct in terms of overall patterns of intergenerational mobility, their analysis and conclusions neglect an important characteristic of the labor force--women and men do not have the same jobs. Therefore, intergenerational mobility may mean something different to women than to men for actual status attainment. Women may move in the same direction as men do but they are moving within two relatively separate structures. Secondly, the examination above of trends in the employment of women suggests that drastic changes in the numbers of women working has recently occurred. The changes may not be due to father's or mother's impact on women, but rather to historical developments in the occupational 29 structure. Intergenerational analysis, then, may be an inadequate way to illustrate the change in women's working lives because it does not account specifically for these historical changes. While it could be used to show how the occupational structure has changed in terms of new demands in occupations, it would be inappropriate here. Studying the movement of one group of women on a micro, status attainment level, might lend insights into the changing structure of opportunities for women where an intergenerational approach would not. The most recent attempt at studying the intragenerational movement of women in the labor force was done by Wendy Wolf and Rachel Rosenfeld. Using a dual labor market approach, Wolf and Rosenfeld proposed the following hypothesis: women's occupations are usually excluded from sectors of the labor market where there are more chances for upward mobility. Using Census data, they define "female occupations" as ones which are composed of 70 to 100 percent female workers. "Male occupations" are 70 to 100 percent men in composition, and mixed occupations are defined as ones with less than 70 percent male or female members (Wolf and Rosenfeld, 1979). Wolf and Rosenfeld found that female occupations offered limited opportunity for movement from one occupation to another occupation. For both men and women, a change from one female job to another female job showed no average change in socioeconomic status. Women in male occupations 30 showed no change in socioeconomic status, although they were less likely to experience downward movement in male jobs than in female jobs. Wolf and Rosenfeld conclude that the greatest change in mobility, either upward or downward, comes from changing job sectors (i.e., male to female or female to male) (Wolf and Rosenfeld, 1978). Wolf and Rosenfeld define primary sector jobs as those which are predominantly male, and secondary sector jobs as those in which females dominate. Their findings are thus tautological-—women's occupations are characterized by women's employment patterns, and men's occupations are defined by their employment patterns. My analysis will remedy this error by identifying primary and secondary occupations independent of their sex composition. I will investigate the distribution of women across sectoral occupations, and examine the consequences of this distribution for single women versus married women. Few studies to date have looked at marital status as an independent variable, which could have an effect on women's intragenerational occupational attainment. Most studies have focused instead on how women's status changes with marriage by measuring the difference between father's occupation and husband's occupation (see, for example, Chase, 1975; Tyree and Treas, 1974). There are, however, two studies which have examined the status attainment of working women while considering marital status, and a look at their findings is in order. 31 Donald J. Treiman and Kermit Terrell used a sample of 2524 employed women aged 30-44 and a sample of 7298 employed men aged 30-44, and compared the process of status attainment for the two groups in terms of educational attainment, occupational prestige, and income level. They controlled for marital status on the income variable only. They showed in their analysis that the process of educational and occupational achievement for women and men was similar. The educational level of parents was the most important determinant of educational level for both men and women. Women and men also came out nearly identical when occupational prestige was measured, and this prestige was seen as contingent upon educational attainment. Educational attainment was found to be the most important determinant of occupational prestige (Treiman and Terrell, 1975). In terms of income level, Treiman and Terrell found dissimilarities between men and women. Women earned substantially less than men, even when hours worked and work experience were controlled. They also found that married women earned far less than never married women when they introduced the same controls, although never married women still earned less than men (Treiman and Terrell, 1975). Treiman and Terrell concluded that the process of status attainment is similar for women and men, although there are discrepancies between women and men, and between married and never married women, when income levels are studied (Treiman and Terrell, 1975). 32 McKee McClendon replicated and extended Treiman and Terrell's research by adding new variables to their model of status attainment. McClendon's sample consisted of 1381 white males and 778 white females, all of whom were 18 years or older and currently employed at the time of the survey (McClendon, 1976). Using Duncan's socioeconomic index (SEI) to measure occupational status, McClendon showed that the overall shape of the distribution of males and females across SEI categories was quite similar, although men were more likely to be in lower status jobs or higher status jobs, and women were not usually found in either extreme. Occupational status was found to be the same for women and men, and education was shown to be the strongest determinant of that occupational status. These findings confirmed those of Treiman and Terrell (McClendon, 1976). McClendon extended Treiman and Terrell's research by adding three new variables to their model of status attainment: marital status, number of children by age group, and work status (e.g., part-time versus full-time). Of the three new variables introduced, he found that work status had the most significant impact on occupational status for women. Number and age group of children did not have any effect on occupational status, and marital status had only a weak effect--currently married women had a 3.5 occupational status point advantage over nonmarried women. McClendon says that this last finding might be explained by 33 the fact that married women have more "freedom" than nonmarried women when choosing whether or not to work. Because they are married, married women are usually not the breadwinners as nonmarried women are. Thus married women can be more choosey when picking a job, and they can refuse to take jobs of low status. In contrast, nonmarried women must take whatever job they can get to live, and they therefore may end up with lower status jobs (McClendon, 1976). Treiman and Terrell, and McClendon posit theoretical models of status attainment. Treiman and Terrell discuss the status attainment process as a constellation of three factors: educational attainment, occupational prestige, and income level. In their model, social origin-~in terms of father's education, mother's education, and father's occupation--is thought to determine educational attainment, which in turn determines occupational prestige and income levels. Treiman and Terrell show that the first two processes operate in a similar manner for women and men, but that the third one is differentiated by sex and marital status. McClendon expands the basic model of status attainment by including three unique factors for women-- marital status, number of children and their ages, and work status. He concludes that the status attainment process is very much the same as Treiman and Terrell had described it, and that his three additional variables did not alter significantly the determinants of women's status attainment 34 (Treiman and Terrell, 1975; McClendon, 1976). The model used in this analysis will extend both of the above models, while also stepping back to include one variable in McClendon's scheme--marital status. Primarily, I want to move past the process of status and/or occupational attainment per se, and focus more on changes in occupational status due to upward or downward mobility. Treiman and Terrell and McClendon agree in their research that the process of occupational status attainment for women and men is the same, and that this process operates mainly through education. Neither of the two studies, however, go on to show what happens to this status after a person has achieved it--under what circumstances it changes or does not change, how it changes, or why it changes. Further, the models generally compare women to men but they do not include as a salient feature possible differences among women of varying socio-demographic characteristics. I want to address each of these deficiencies in some part in the following analysis. More specifically, I want to investigate the changes in occupational status over time for a select group of women, while showing what impact, if any, marital status has on such occupational mobility. In order to do this, I will draw upon insights from dual labor market theory, and use them as a predictive framework for women's mobility. Dual labor market theory contends that there are two kinds of jobs: primary and secondary. Fundamental 35 differences between the two exist, one of which is differential chances for upward mobility. Because primary jobs are characterized by highly developed internal labor markets which have defined job progressions, they are likely to offer a worker some chances for upward mobility. Secondary jobs, by distinction, have relatively undeveloped internal labor markets with no defined job progressions, and are therefore less likely to present a worker with opportunities for upward mobility. This analysis will use this distinction to test the relationship between women's marital status and their allocation to primary or secondary jobs and their occupational mobility. Women who are married generally are not breadwinners. Their participation in the labor market is contingent upon family needs, and they move in and out of the labor force to adapt to these needs. Their participation in paid wage labor may not be for long periods of time. This sporadic employment pattern indicates that married women would be expected to be allocated to secondary jobs, which are characterized by high worker turnover rates. Location in these secondary jobs would have severe consequences for these married women in terms of upward mobility. In contrast, women who are single are more likely to be their own breadwinners. Their need for continual employment is generally greater than it would be for married women. Since their participation in the labor force is central to their survival, single women would be expected to be in 36 primary jobs, where job benefits and wages are higher than in secondary jobs. Location in primary jobs would give single women access to job progressions and internal labor markets, and they would be upwardly mobile. If marital status is not significant in predicting allocation to kind of job, it is thought that kind of job will determine rates of directionality of movement nonetheless. This is because being in a primary job or a secondary job should have significant consequences for mobility, as dual labor market theory suggests. The analysis which follows will thus test four hypotheses in order to determine if there is any evidence to support the claims made above. The four hypotheses are: 1. Single women are more likely than married women to have had an initial occupation in a primary job. 2. Single women are more likely than married women to have had upward mobility between first job and a later job. 3. Women whose initial jobs are primary are more likely to be upwardly mobile than women whose initial jobs are secondary. 4. After category (primary/secondary) of first job has been controlled, the differences in extent of upward mobility between married women and single women should disappear. After a brief description of the research design and data sets to be used, the four hypotheses will be tested. CHAPTER III RESEARCH DESIGN AND DATA DESCRIPTIONS This project will use two different data sources to show what relationship, if any, a woman's marital status has to her chances for intragenerational mobility. Two samples are needed to accomplish this, as will become evident below. We have already seen that, theoretically, chances for mobility are important determinants of whether or not an occupation can be labelled primary or secondary. The first task in the analysis will thus be to rank a list of Census occupation codes according to their mobility potential. To do this, I will use the Occupational Changes in a Generation-Replicate Master File (OCG-RMF) data set. It was collected by the Census Bureau in 1962 and 1973, as supplements to the 1962 and 1973 March Current Population Surveys. The target population in 1962 was males from 20 to 64 years old in the civilian noninstitutionalized population. The target population in 1973 was males from 20 to 65 years old in the civilian noninstitutionalized population. The latter study was a replication and an extension of the 1962 OCG survey. Selected data items from the two dates were merged to permit examination of changes 37 38 in social mobility in the period between the early 19605 and the 19705. The data include education, occupation, industry, and earnings for men and their wives; number of siblings and paternal education, occupation, and industry for men and their wives; and additional social background and work history variables for men. The combined samples include a total Of 62651 respondents. Using this data set, then, Census occupations will be divided into five categories which are stratified by their potential Opportunity for mobility--high movement upward, low movement upward, no movement, low movement downward, and high movement downward. Those occupations in the first two categories will be defined as "primary" jobs, and occupations in the last three categories will be labelled as "secondary" jobs. Potential mobility is calculated by subtracting the Duncan socioeconomic index in 1962 from the average Duncan socioeconomic index in 1973 for the 1962 occupants of each of the 297 Census occupation categories. After this classification is obtained and each Census occupation category has been designated as either primary or secondary, I will use a sample of females to determine their actual mobility over time. It is important to point out the reason why the initial sample used to designate occupations as either primary or secondary in mobility potential is one that is all male and not female. There are simply no samples available which contain longitudinal information on women's labor force 39 experience that covers as broad an age span and as long an interval as this sample does. Using the Occupational Changes in a Generation-Replicate Master File data will allow estimations based on men's experience in all age cohorts, and will give a more accurate picture of structural opportunities for movement than a more limited sample provides. The second data set to be used will be that which was collected by the Census Bureau for the Center for Human Resource Research on women's labor market experience (the Parnes data). The sample is a probability sample of civilian, noninstitutionalized women in the United States who were 30-44 years old in 1967. Data was obtained through personal interviews. The survey was reconducted periodically after 1967, and this analysis will use as its follow-up time 1974. In each survey, detailed information was gathered on labor force and employment status, and on labor market experience and income, in order to analyze the relationships over time between labor force experiences and other social and economic characteristics. The total number of women in this sample is 5083. In order to look directly at marital status as an independent variable, I have deleted from the Parnes sample those respondents who were married more than once, and those widowed, divorced, or separated. I have also deleted women who did not work continuously between 1967 and 1974, and women whose marital status changed from never married in 40 1967 to married in 1974. Using women who worked continuously between 1967 and 1974 should provide for a better test of mobility, since length of time in the labor force is thought to be critical to chances for mobility. And the only time controls possible with this data set were those between 1967 and 1974. Women who married for the first time between 1967 and 1974 were eliminated because their numbers were too small to be significant, and their inclusion may have, in fact, confounded the impact of marital status. I have also not included women of color in this study because the factors of race and racism are ones which complicate and compound an already complex issue. A broader, more complete study of women's occupational mobility would, of course, take all these variables into consideration. The sub-sample I will deal with thus consists of 1691 women who were married only once and whose spouse was present in 1974, and 103 women who were never married. Table 3 shows a complete sequential breakdown of how and in what what quantity women were deleted from the sample. This analysis of women's data will involve two stages. Primarily, I will determine whether or not marital status has any relationship to a woman's location in either a primary or a secondary occupation. I will look at this distribution of women across types of occupations for first occupation ever held after full-time schooling had stopped. Looking at first occupation will be important since dual 41 labor market theory suggests that the occupation a person initially enters will seriously affect later chances for movement. And it should then be evident from the first stage of this analysis whether or not a particular job classification has some mobility ladder attached to it. Table 3: WOMEN EXCLUDED FROM SAMPLE FOR CURRENT ANALYSIS Number Excluded Reason for Exclusion 1627 Black women 311 Widowed, divorced, or separated women 377 Women married more than once 956 Women not working in labor force continuously between 1967 and 1974 18 Women for whom marital status changed from never married in 1967 to married in 1974 3289 Total Women Eliminated 1794 Total Women Retained 5083 Total Sample Secondly, I will examine movement in occupations from first job to 1974 job, again analyzing what relation marital status has to that change. That is, I will look at women who were married only once in 1974 to see what proportion moved up, what percentage made no move, and what proportion moved down. I will repeat this process of women who never married. I would expect that married women are more likely 42 found in secondary occupations when they initially entered the labor force and therefore would have downward or stable mobility in 1974. By contrast, single women should be found in higher proportions in primary occupations at the entry level and thus should have higher rates of upward mobility when compared to married women in 1974. What then, is the impact of marital status on: 1) women's location in primary or secondary occupations, and; 2) women's opportunities for mobility? The following analysis should give some answers. CHAPTER IV DATA ANALYSIS The first step in this analysis is to use the Occupational Changes in a Generation-Replicate Master File to arrive at five categories of occupations which are stratified by their potentials for mobility. This is done by subtracting socioeconomic index score (SEI) in 1962 from SEI score in 1973 for the 1962 occupants of each Census occupation. The five mobility categories were defined in the following way: ”high movement upward” occupations were those whose average SEI difference from 1962 to 1973 was +15 points or greater; "low movement upward" showed an average SEI difference of +5 to +14.9; I'no movement" occupations had an average SEI difference Of +4.9 to -4.9; "low movement downward," -5 to -l4.9 average SEI difference; and, "high movement downward," -15 or less average SEI difference. The Appendix gives a complete listing of Census occupations which fall into each of the above five categories. The second stage of this research project will test the four hypotheses listed above, using the data set on women aged 30-44 in 1967. Hypothesis 1 predicts that single women are more likely 43 44 than married women to have had an initial occupation in a primary job. Table 4 provides a test for this assertion by crosstabulating marital status with the mobility potential of first occupation. Never married women are more likely than married women to be located in an initial occupation which has high upward mobility potential. Married women, however, have a greater likelihood than never married women of being in a low upward mobility potential occupation at first job, and this difference is quite substantial. Never married women are also more likely than married women to be in secondary occupations with respect to mobility potential when they enter the labor force. This pattern between married women and single women, where married women are less likely than single women to be in either extreme of mobility potential, mirrors precisely the pattern observed by McClendon between women and men. This may indicate that McClendon's analysis is correct--married women can be more selective than single women when choosing an occupation, since they are not usually the main breadwinners in a marital dyad. They may choose a job with slightly higher status in terms of mobility potential. Single women, on the other hand, may have to take a job with lower mobility potential in order to support themselves. Single women are also, however, represented in higher proportions in the highly mobile upward category, which may point to their need for continual steady employment and jobs that offer opportunities for much upward mobility (McClendon, 1976). 45 Ho>ma mo. um unmofimwcmflm mumsvm ago wo.v mm.m wm.mm mm.Hm w¢.HH embanz mm" own: ommus ommu: mom": mn.m mm.¢ mm.h wo.mm mm.mm wm.ma moans mu: mu: mmuc Hens man: wm.¢m wm.¢ mH.m wa.mm mo.~m wm.HH Hmoau mhu mmu mmvn mnmu mmHu amuoa 5309 :309 no mo 30m w>oz swam O>Oz 30A o>oz Oz m>oz 30a O>Oz :mHm Hmuoe cEnHOU vwwuumz um>oz pmwuumz mDB¢Bm AfiBHmOH Ho. um ucmowmecmflm mumsvm gnu wm.mH wa.oa ws.ms wo.HH w~.mfl Hobos «mesanz msmuc means opens smauc mmNuc assaou wn.m wm.m mm.m wa.ms mm.oa wm.mm museum: NOHHC QHC mHC Dfiflc HHHC FNHC H0>wz wm.am wo.o~ mm.oa wm.ms wH.HH wm.ma «swans smmu onau amen mmauc Hem" smashes amuoe c303 2309 no as 30m w>oz nmflm m>oz so; m>oz oz w>oz 30; m>oz swam maeasm SaaHmwa Hoo. um DCMOfiwflcmfim mumsvm “no mo.om wm.oa m~.ma wa.HH wo.mH Houos 4seofiuz 4mm" aha": amen: mmauc How": csofioo wn.s wo.om mm.HH wa.~m wa.m wm.H mun mmnc mu mm": an: H": 2300 m>oz swam wo.m wm.as wh.mH mm.mm mm.m mo.o Hm" HN" w" “H“ m" on C309 m>02 30q wa.m~ mm.o~ w~.m mm.m¢ wo.m wo.n swan: mmauc osuc menu mm" emu o>oz oz wo.~m wo.¢a wm.OH WH.H¢ wm.mH wo.o~ Hem" swan emu mmmn mHHu «pan on o>oz zoq mm.HH »~.0H wN.OH mm.ms wa.oa w~.om swan man man Hm" man as" on o>oz gown souoe czoo csoo on on onaoz zmfiz oooz 3oz o>oz oz o>oz 3oz o>o2 swam aaHezmeom squHmoz ZmEOZ DmHmm¢2 mom ZOHBOH Hoo. um DCMOflchmHm mumswm Hnu wm.m mm.m wa.me wa.oa wm.m~ Hobos *NOHu mu mac we": Han: emu csoaou wm.v wo.ov wo.o mo.oe wo.o wo.om mus «us on: mu: on H": csoa O>Oz Loam mm.s wo.o wm.~H wm.mm wo.m~ wo.o m" C" HHC m" N" O" 830nm 0>Oz 30..”— w¢.mm w~.ma ma.m wo.mo mo.m wH.m mm": mu mu Hm" an mu o>oz oz wm.o¢ mm.¢ m~.~H wo.om wm.ma mm.om Hen: mu mu: man: mu Han: ms o>oz son wn.ma mo.o wo.o wo.o~ wo.o wo.om man on one mu on: «an: as O>Oz swam amuos czoa czoa as a: oneoz swam o>oz zoq o>oz oz o>oz zoo o>oz omen Omz mom ZOHedeUUO vhmH OE ZOHeflmDUUO EmmHm 20mm wfiHQHmOZ AflDBU< Mm AfiHBZmBOm MBHquOE ZOHaflmDUUO EmmHm uh OHQMB 51 move. Women whose initial jobs were low but upwardly mobile in potential were more likely to make a high move up than a low move up. Women whose initial jobs were no moving, and low or high downwardly moving were less likely to experience upward mobility. Mobility potential of first occupation, irrespective of marital status, is a strong determinant of actual mobility. Hypothesis 3 is substantiated by the data. Tables 6 and 7 also provide a test for Hypothesis 3 while controlling for marital status. In general, and somewhat irrespective of first occupation mobility potential, single women were more likely to move up or make no move down than married women. Again, being single protects a woman from downward mobility and insures some upward mobility, particularly when compared to women who are married. This discussion, however, should not ignore the fact that a large percentage of women in all categories of first occupation mobility potential did not move. The most likely outcome for four of the five categories of initial occupation is no move (less than five points in either direction); and one should note that in the one deviant category, the most likely movement is high downward. This study has not been designed in a way that would explore why this is so, but two possible explanations can be proposed. Either women do not move because of direct discrimination against women, or women are unable to move because the structure of occupational Opportunity is such that only a 52 small percentage of any group of people would move. Ultimately, answers to this problem could be gotten by comparing these women with a similar sample of men. Since there is no comparative base for examining why this is so, the pattern can only be recognized here and perhaps investigated at some other time. The final hypothesis to be tested here is a crucial one, for it asserts that the differences in extent of upward mobility between married women and single women shall disappear after category of first job has been controlled. This hypothesis tries to ascertain which variable--"marital status" or "mobility potential of first occupation" (as primary or secondary)--is more important in determining actual mobility. If the former is more important, the tables should each show a strong relationship in the expected direction; if the latter is more important, the tables should show no significant relationships. Tables 8 thru 12 show the relationship between marital status and actual mobility between first job and 1974 occupation while controlling for each category of potential mobility at first occupation. The tables indicate that marital status is most important for actual mobility in the mobility potential occupations which are highly mobile upward. Looking at Table 8, single women are much more likely to make an actual move which is high and upward than married women, while married women are more likely to move low and upward, to not move, or to move downward, either high or low, than single 53 HO>OH Hoo. um unmowmacmflm mumswm ecu w¢.m w¢.m mm.H¢ wv.m wm.om Hmuoa mom" man: man: vmu man: Hm": assaou we.» $0.0 mo.o wo.om wo.o wo.ow peanumz man: on: on: mu on: NHHC uw>wz mm.mm wm.oa mm.oa mm.m¢ wm.oa wm.w~ swan man man Hmuc man man coauumz qmuoe csoo ssoa no mo 30m O>Oz swam O>Oz 304 o>oz Oz o>oz 304 O>Oz swam mDBoz nose. mH SaHszmaom squHmoz oneoz pm.mm wm.¢H mm.oa wH.Hv wm.ma wo.om Hhmu hmau «mu: wmmu mafia: «ban mowuumz amuoa 2309 c309 a: a: 30m O>Oz swam O>Oz 30a o>oz oz o>oz so; o>oz swam moassm q¢BHmoz sod. mH anszmsom squHmoz one.s we.» Hobos ommuz mmauc mane sown: own: osuc esoHoO mm.o mm.ma wa.m mo.mo mo.m wa.m ooflouoz mm" mu mus Hm" an: mu: oo>oz mn.mm mm.o~ wm.m mm.mv wo.m mo.h has" mafia own: menu: mm": emu: oofluooz amuoe csoo axon as m: 3oz o>oz zmflz o>oz sou o>oz oz oooz zoo o>oz zmflz mesmem qoz oz: mH AdHazmeom MBHAHmOZ ZOHadeUUO BmMHm mmomz szOB mmOmB mom .mDEOH OH. um ucmowMHcmflm mumsvm «no wo.mm wm.ma wm.sm mm.HH wo.o Houos mm" Hm": on «an: en on: assaoo wo.ma wo.o wm.ma mm.~o wo.m~ wo.o oofluuoz w“ ONC H" mug N" ONE Hw>wz ms.om w~.Ha wh.ma mm.mm wm.m wo.o Hm" Hm" mu 5H" mus on: oofiuooz qmuoe csoo c300 as as 3oz o>oz swam o>oz 3oz o>oz oz o>oz zoo o>oz amen massam qasHmaz :c3oa w>oz 304: mH AdHBzmeom MBHAHmOS ZOHBoa OH. um unmoflmwcmfim mumsvm Hso ww.mv ww.oa wm.mm wm.v we.~ Hmuoe *mw" H?" QHC FN" @"C NHC CEDHOU wo.m mo.ov wo.o wo.oe wo.o wo.o~ pofluuoz mus N" OHS N" OHC H" Hm>mz wo.¢m $0.0m wm.aa wa.mm wH.m mm.H mun mm": mu man: an: an: cmfiuumz qmuoa 5300 c300 no as 3cm o>oz swam o>oz so; o>oz oz o>oz so; o>oz swam moeaem gasHmoz swam. mH SOH moo. um ucmOflmacmflm mumsvm “no wm.ma wo.oa wo.as wo.m~ wa.- Hobos sHHHu mean: maaue ems": mean: mam": esofloo wo.m wo.m wm.m mH.Nm wm.sa ma.as oofloooz wmflc NHC MHC wHHC mug MN" Hw>wz wo.mm wm.ma ws.oa wm.fls wm.ma wH.H~ mmoan: means maau amen smauc mum" oofluuoz sauce csoa c309 no mo 3oz o>oz zmflz o>oz zoo o>oz oz o>oz 3oq o>oz nose mzaasm qaeHmaz gzuosflomg mH qw wn.> wm.m kuoa emmmu mafia: Hon: mam": Hmus men: sasaou wm.m wm.ma m>.m wm.om wm.m wn.m coauumz own bu was mun: mu: en: um>oz wa.mm wm.om mm.m wm.me mm.> wm.m mam" mmau hm" mam": mun mm" omwuumz amuoe :300 :300 an a: 30m w>oz swam O>Oz 304 o>oz Oz O>Oz soq O>Oz swam mnemam adaHm