|
|
A Longitudinal Analysis Of College-Educated Women's Self-Employment Decisions by Barbara Woods McElroy Paula D. Harveston Cherlyn S. Granrose
|
|
|
A Longitudinal Analysis Of College-Educated Women's Self-Employment Decisions by Barbara Woods McElroy Paula D. Harveston Cherlyn S. Granrose
|
Barbara Woods McElroy is a professor at Susquehanna University's Sigmund
Weis School of Business. She can be contacted at mcelroyb(at)susqu.edu by
replacing (at) with @. Paula D. Harveston and Cherlyn S. Granrose are both
professors at Berry College's Campbell School of Business.
|
ABSTRACT Researchers do not agree on the reasons why women become entrepreneurs. In this article we use longitudinal data collected over a twenty-year period to explore whether college-educated women entrepreneurs were different early in their lives from other women. Early differences would offer evidence that female entrepreneurs, much like their male counterparts, are led to self-employment by its inherent virtues or by their specific talents or interests. Lack of early differences would lend credence to the notion that female entrepreneurs are led to self-employment by other factors. We also examined the responses of the entrepreneurial women studied more qualitatively in an attempt to categorize the entrepreneurs into the four categories recommended by Goffee and Scase (1985). Based upon a longitudinal analysis, only about half of the women seemed to fit into these categories, and the number who seemed to fit changes at each data point. The evidence suggests that women’s status as entrepreneurs is at least as much the result of life circumstances as the completion of long-term goals. |
 |
INTRODUCTION |
Over the past decade, an increasing number of women have started businesses. A number of studies have considered several aspects of female entrepreneurs: their activities (i.e. Buttner, 2001); career selection (Matthews and Moser, 1996; Scherer, Brodzinski and Weibe, 1990); work-home role conflict (Stoner, Hartman and Arora, 1990); gender and ownership patterns (Rosa and Hamilton, 1994); entrepreneurial attitudes and skills (Sexton and Bowman-Upton, 1990); perceptions about entrepreneurship (Hisrich, Koiranen and Hyrski, 1996); education (Dolinsky, Caputo, Parsumarty and Quazi, 1993; Hisrich and Brush, 1983); and networking patterns (Aldrich, Reese and Dubini, 1989; Andre, 1992; Carsrud, Gaglio and Olin, 1986; Cromie and Birley, 1992). The majority of these studies compared male and female entrepreneurs. While these studies show the differences between female and male entrepreneurs, collectively they fail to clarify the unique dimensions of women entrepreneurs.
The focus of this study is to examine over time why college-educated women choose to become self-employed entrepreneurs. We are particularly interested in examining whether long term intentions drive entrepreneurship for women or whether they are forced into self-employment by lack of opportunity due to the existence of a glass ceiling in the corporate world or are led to it by opportunities that present themselves as their lives progress (Auster, 1993). We investigate this research question using longitudinal data gathered from a sample of women we followed over a twenty-year period after their formative college years.
|
|
THEORETICAL BACKGROUND |
The management literature contains many studies of entrepreneurs. Entrepreneurship is important because small businesses have historically provided both the majority of jobs and innovation in the U. S. economy. Recently, this trend has intensified and become more global in nature. As shown in Table 1, the percentage of jobs in firms with more than one hundred employees has fallen in nearly all developed nations (Reynolds and White, 1997).
Table 1
|
Country |
Years
|
Firm Size (# of employees, %) |
|||
|
|
|
1-19 |
20-99 |
100-499 |
500+ |
|
Canada1 |
1984-89 |
0.6 |
6.8 |
-2.1 |
-3.7 |
|
France2 |
1985-90 |
3.0 |
0.0 |
-1.9 |
-1.1 |
|
Germany |
1988-90 |
2.1 |
-0.7 |
-0.4 |
-0.9 |
|
Italy |
1988-90 |
3.6 |
-0.8 |
-0.3 |
-2.0 |
|
Japan2 |
1988-91 |
-0.4 |
0.4 |
0.2 |
-0.7 |
|
Spain |
1988-90 |
7.3 |
-3.8 |
-4.5 |
1.0 |
|
UK |
1988-91 |
1.0 |
1.0 |
-0.5 |
-1.5 |
|
USA3 |
1988-90 |
7.4 |
-1.1 |
-1.5 |
-3.2 |
Adapted from Reynolds and
White, 1997
1Manufacturing
only, 2Employees only (Owners excluded), 3Establishments,
excluding agricultural sector
Generally, early entrepreneurship research was conducted by economists who focused on the role of the entrepreneur in economic growth and innovation (i.e., Schumpeter, 1934). Early studies found that new businesses were the result of a desire to be entrepreneurs, and that entrepreneurs were somehow different from employees (Brockhaus, 1982; Brush, 1992).
Beginning in the 1970s, a second wave of interest in entrepreneurship emerged that compares female and male entrepreneurs. This surge of interest in female entrepreneurs was fueled by lackluster performance of major corporations, by the women’s movement, and by the fact that female self-employment growth outpaced that of males. Female self-employment in the UK increased 81% between 1981 and 1989, while male self-employment increased only 51% (Carter and Cannon, 1992). In the U.S., women started businesses at twice the rate of men between 1975 and 1990 (Murphy, 1992).
This second wave of research found that female entrepreneurs differ from their male counterparts in several ways. Women attach different connotations to terms commonly used to describe entrepreneurs such as “innovation” (Lee-Gosselin and Grise, 1990) and “risk taking” (Green and Cohen, 1995). Hakim (1989) found that women become entrepreneurs for different reasons than men. Early research using male subjects indicated that entrepreneurship was the result of a lifelong desire for self-employment, and that entrepreneurs and employees were inherently different. More recent research with women indicated that entrepreneurship could also be a response to dissatisfaction with career options (Rosin and Koranik, 1992). Women entrepreneurs appear to vary greatly from their male counterparts in the resources such as spousal support, education, and work experience that they bring to the new enterprise (Watkins and Watkins, 1984; Harveston, Davis and Lyden, 1997). Finally, there is evidence that women organize their business operations and manage succession issues differently (Harveston, Davis and Lyden, 1997).
Perhaps Green and Cohen (1995) best described the effect of these differences. Their grounded-theory study explored the experiences of twenty four women who left large corporations to set up their own businesses. They concluded “while self-employment offers ways of accommodating women’s dual roles as mothers and professionals, it does little to alter their structural positions within the labor market and society more generally, nor does it seem to have any impact on their ideological perspectives. Self-employment in this sense can be seen as a hegemonic process–giving women the opportunity to negotiate, to feel autonomous, empowered, and in control, and in doing so, ensure that existing circumstances remain essentially unchallenged.”
Review of these differences in the past literature led us to the question of whether women who become self-employed differ early in their lives from those who do not become entrepreneurs. Early differences between those who become entrepreneurs and others would support the idea that female entrepreneurs, much like their male counterparts, are led to self-employment by its inherent virtues or by their specific talents or interests. Lack of such differences would lend credence to the notion that females become entrepreneurs due to their dissatisfaction with opportunities available to them as employees or opportunities that arise later in their careers or because of opportunities that the marketplace affords them. We wanted to examine several potential areas of difference between entrepreneurs and employees. Thus, the primary research questions we pursued were whether women who become entrepreneurs are identifiable by characteristics of family background, by their career orientation, by their beliefs about family relationships, by their expectations about family income, by their feelings of self-worth, by their levels of satisfaction with family relationships or their careers. Therefore, we decided to test the following hypotheses:
H1: Entrepreneurs and employees will have different views about family relationships.
H2: Entrepreneurs and employees will have different career orientations.
H3: Entrepreneurs and employees will have different expectations about family income.
H4: Entrepreneurs and employees will have different views of their self worth.
H5: Entrepreneurs and employees will have different satisfaction levels with their family relationships.
H6: Entrepreneurs and employees will have different satisfaction levels with their careers.
|
|
EXISTING CLARIFICATIONS OF FEMALE ENTREPRENEURS |
In addition to testing for differences between female entrepreneurs and employees we wanted to explore differences among female entrepreneurs. We considered two classification schemes from extant literature. Goffee and Scase (1985) classified women entrepreneurs into four types based upon two factors: their attachment to entrepreneurial ideals and their willingness to accept conventional gender roles. Conventionals showed commitment to both factors. They were primarily working-class women with fragmented work histories forced into self-employment by economic necessity. Innovators were committed to entrepreneurial ideals but not to conventional gender roles.They were largely professional women who had chosen self-employment as a solution to restricted career prospects. Domestic Entrepreneurs had little attachment to entrepreneurial ideals and much to conventional gender roles. They tended to fit their business duties around their family responsibilities. Radical Entrepreneurs held little attachment to either factor. They were frequently involved in collective ventures, political in nature, and aiming to promote women’s interests.
Figure 1
Goffee and Scase Typology of Female Entrepreneurs
| Commitment to Entrepreneurial Ideals | |||
| Commitment to | Low | High | |
| Traditional | Low | Radical Entrepreneurs | Innovators |
| Gender Roles | High | Domestic Entrepreneurs | Conventionals |
From Goffee and Scase (1985)
Carter and Cannon (1992) use a sample of 60 business owners in Britain to classify female entrepreneurs into five types based upon heterogeneity and change. Drifters are young women who use self-employment to overcome unemployment.Young achievers are aspiring, well educated and with little experience. Achievers were somewhat older and with more experience. Returners used self-employment as a way to re-enter the work force after a career break. Traditionalists were women, generally over age 45, who had always worked in family- owned and managed businesses. Carter and Cannon stress the flexibility of these categories, and the dynamic nature of small business.
Each of these classification schemes has limitations. The Goffee and Scase typology is largely theoretical, and it represents a snapshot of entrepreneurs at a given moment. The Carter and Cannon taxonomy better reflects the changing nature of entrepreneurship as women pass through the stages of life. We chose to use the Goffee and Scase categorization for two reasons. First, it addresses women’s motivations for self-employment more effectively. Second, our sample does not include older women, and that eliminates the possibility of having any subjects in Carter and Cannon’s final two categories.
|
|
METHOD |
In 1980-82, 433 college junior and senior women participated in a study of the intention to return to work following childbirth. Respondents were asked to describe these intentions, their beliefs about consequences of working and not working, and their career and childbearing plans.
In 1990, we obtained 336 current addresses of the original participants. We received Phase II questionnaires from 263 women, and were able to match Phase I and Phase II questionnaires for 228 women (68% of those sent questionnaires and 52.6% of the original respondents). In 1990, two additional types of questions were asked. The first asked the extent to which job characteristics, family situations, and child care problems would influence the person to leave an organization. Respondents were asked to refer to the organization they were working for at the time of the birth of their first child if they were mothers, or to their current organization if they were not mothers. A second type asked the extent to which their expectations had been met in the jobs they had held since college and whether a woman had ever considered becoming an entrepreneur
In the summer of 2001 we located for the third time 265 women. Of those, 75 responses could be matched to both Phase I and Phase II (29% of those sent Phase III questionnaires and 17% of the original participants). Though this would generally be considered a rather low response rate, we were pleased to realize that 29% of the women we located retained an interest in the study twenty years after its inception despite lives busy with work and children. In addition to repeating the questions from Phase One and Phase Two about respondents’ beliefs about the consequences of working or not working after childbirth, the 2001 survey included questions about whether the woman had considered self-employment, whether she actually opened the business and, if not, why not, whether either of her parents had been an entrepreneur, and the response of any partner to her desire to become self-employed.
|
|
DEMOGRAPHIC INFORMATION |
Phase I:
As reported earlier, the original 433 women participating in the study were college juniors and seniors.They came from one of two universities in the northeast and were pursuing a wide range of majors.
Phase II:
As a group, the women who participated in Phase II of this study maintained an educated but middle class life style during the 1980s. Over one-third received graduate degrees. The majority (57%) was employed in service industries, with substantial minorities in finance, insurance and real estate (20%), and manufacturing (11%). The most common occupations included manager (25%), housewife (11%), scientist (9%), and computer specialist (8%). Over 72% worked 35 or more hours per week in jobs they had held for the past three years. Their average income was slightly over $30,000, and for those who were married, spouses made about $50,000 annually. Family life had been delayed longer than anticipated in 1980, but 46% had one child, 20% had two children and 4% had three.
Phase III:
By the summer of 2001, only one woman continued to delay childbearing. The respondents are on the whole middle-class. Because some women who self-reported as entrepreneurs reported very few hours of work, we classified women who work fewer than ten hours weekly as homemakers in order not to confuse entrepreneurs with those who only had an avocation. Of the 75 current respondents, we classified twenty as entrepreneurs, 42 as employees, and 13 as homemakers.[1]
|
|
RESULTS |
Demographic differences
The only significant demographic differences among the three groups were that entrepreneurs had a larger number of younger brothers, and that a larger portion of entrepreneurs had a self-employed parent.
Analysis
We conducted two types of analysis. The first analysis was simple F-tests to discover mean differences between the three groups on a variety of measures gathered as college students and at one and two decades after college graduation. All items were measured on a 1-5 scale with 1 indicating events of very low value or very improbable events and 5 indicating very valuable or very probable events. The second analysis was a qualitative analysis of the responses the entrepreneurial women made about their careers at all three times in an attempt to categorize the respondents into the Goffee and Scase model.
The Statistical Analysis
We compared the scores of all women who returned questionnaires for Phase Two and Phase Three on these questions about their interest in becoming an entrepreneur. Because we have lost track of many participants in the past twenty years, this is a small sample. Many differences were non-significant. However, a few were significant in ways that are interesting.
H1: Entrepreneurs and employees will have different views about family relationships
Table 2A: Family Relationships
College Years
|
Variable
Description |
Mean Entrepreneurs N=20 |
Mean Employees N=42 |
Mean Homemakers N=13 |
Significance |
|
Values: |
|
|
|
|
|
Feel close to child |
4.71 |
5.00 |
5.00 |
** |
|
Train child myself |
4.00 |
3.74 |
4.14 |
|
|
Have time for Child |
4.36 |
4.79 |
4.71 |
|
|
Independent child |
4.21 |
4.33 |
4.00 |
|
|
Secure child |
4.71 |
4.78 |
4.86 |
|
|
Well-disciplined child |
4.43 |
4.61 |
4.29 |
|
|
Probability if working: |
|
|
|
|
|
Feel close to child |
3.85 |
3.89 |
3.43 |
. |
|
Have time for Child |
3.54 |
3.63 |
3.00 |
|
|
Independent child |
4.08 |
4.16 |
3.86 |
|
|
Secure child |
3.69 |
4.11 |
3.43 |
|
|
Well-disciplined child |
3.54 |
4.11 |
3.29 |
. |
|
Probability if not working: |
|
|
|
|
|
Feel close to child |
4.08 |
4.53 |
4.14 |
|
|
Train child myself |
4.69 |
4.68 |
4.00 |
. |
|
Have time for Child |
4.69 |
4.89 |
4.43 |
** |
|
Independent child |
3.08 |
3.68 |
3.43 |
|
|
Secure child |
4.38 |
4.53 |
4.14 |
|
|
Well-disciplined child |
4.23 |
4.37 |
4.14 |
|
|
Probability: |
|
|
|
|
|
Partner Approve of working |
4.00 |
4.21 |
3.86 |
** |
|
Comply with Partner’s wishes |
4.00 |
4.21 |
3.86 |
|
|
Parents Approve of working |
0.92 |
1.37 |
0.57 |
|
|
Comply with Parents’ wishes |
0.77 |
1.42 |
0.71 |
. |
* .05 < p < .10
** .01 < p < .05
*** .00 < p< .01
All scales are from 1
(not at all, or not at all likely, or not at all probably) to 5 (very much
or very likely, or very probable)
Table 2A shows the results of questions about family relationships given by the women while of college age. There were very few significant differences, and in two of the three cases those differences were between homemakers and employees rather than between entrepreneurs and employees. The only significant difference in attitudes about family relationships between future entrepreneurs and employees was that entrepreneurs placed less value on feeling close to their children. In light of there being no differences in the value they place on having time for the child, training the child personally, and the child being independent, secure, and well-disciplined, this one difference seems inconsequential. Thus, there is little evidence that entrepreneurs and employees have different attitudes toward family relationships while in college.
One-Decade Post-College Perceptions
Table 2B shows the results of the family relationships questions one decade after college. For many of these women, motherhood had been postponed longer than anticipated. Though 91% expected in 1980 to have children a decade later, only about 50% of subjects had children eight to ten years after college graduation. Though homemakers exhibited some significant differences from the others, there were no differences in the probabilities entrepreneurs and employees assigned to the consequences of working or not working on their family relationships. There is no evidence that entrepreneurs and employees have different attitudes toward family relationships one decade past college.
|
Variable
Description |
Mean Entrepreneurs N=20 |
Mean Employees N=42 |
Mean Homemakers N=13 |
Significance |
|
Consequences of working: |
|
|
|
|
|
Feel close to child |
4.77 |
5.63 |
3.71 |
|
|
Train child myself |
4.08 |
3.95 |
4.00 |
|
|
Have time for Child |
4.85 |
4.68 |
4.43 |
|
|
Independent child |
4.46 |
4.53 |
4.43 |
|
|
Secure child |
4.85 |
4.95 |
4.57 |
* |
|
Well-disciplined child |
4.69 |
4.53 |
4.29 |
|
|
Consequences of not working: |
|
|
|
|
|
Feel close to child |
3.92 |
3.84 |
3.29 |
|
|
Train child myself |
2.62 |
2.95 |
2.57 |
|
|
Have time for Child |
3.15 |
3.32 |
2.57 |
|
|
Independent child |
3.62 |
4.21 |
3.57 |
* |
|
Secure child |
3.69 |
3.89 |
3.14 |
|
|
Well-disciplined child |
3.54 |
3.63 |
3.43 |
|
|
Probability: |
|
|
|
|
|
Partner Approve of working |
4.00 |
4.11 |
5.29 |
|
|
Comply with Partner’s wishes |
1.92 |
2.58 |
4.29 |
* |
|
Boss Approve of working |
4.85 |
4.37 |
5.14 |
|
|
Comply with Boss’s wishes |
3.38 |
3.74 |
4.43 |
|
* .05 < p < .10
** .01 < p < .05
*** .00 < p< .01
All scales are from 1
(not at all, or not at all likely, or not at all probably) to 5 (very much
or very likely, or very probable)
Two Decades Post-College Perceptions
Table 2C presents the results of the family relationships questions two decades post-college. By this point, all but one of our respondents had children. We found only one significant difference between entrepreneurs and employees. Employees are more likely to believe that their child will be secure if they work. Perhaps the fact that homemakers and entrepreneurs do not share this explains their choices. Homemakers may stay home in an attempt to make their children secure. Entrepreneurs may choose self-employment because they think that the greater control over their work and their working hours makes their children more secure.
Table 2C: Family Relationships
Twenty Years after College
|
Variable
Description |
Mean Entrepreneurs N=20 |
Mean Employees N=42 |
Mean Homemakers N=13 |
Significance |
|
Consequence of working: |
|
|
