Original Article Gendered Institutions and Cross-National Patterns of Business Creation for Men and Women Amanda Elama and Siri Terjesenb a
Babson College, Wellesley, MA Kelley School of Business, Indiana University, Bloomington, IN
b
Abstract In this article, we explore how gendered entrepreneurship rates are affected by both soft (values, beliefs and expectations) and hard (institutionalized norms and practices) measures of cultural institutions. We use data from the 2001 Global Entrepreneurship Monitor for 25 265 individuals in 11 countries to examine how institutional arrangements related to women’s employment (role of occupational segregation, gender wage inequality, female business leadership and public childcare support) interact with individual-level perceptions in ways that increase women’s start-up. Controlling for national variations in opportunity structure, our results show that gendered institutions (female business leadership, gender wage inequality and public expenditures on childcare) influence the decision to start a business indirectly through perceptions and gender. Dans cet article, nous examinons comment les taux d’entreprenariat fe´minin sont influence´s par les institutions culturelles tant ‘douces’ (valeurs, croyances, attentes) que ‘dures’ (normes et pratiques institutionnalise´es). Nous utilisons des donne´es tire´es de l’enqueˆte Global Entrepreneurship Monitor de 2001 conduite sur 25 265 individus dans 11 pays diffe´rents afin de de´terminer comment les arrangements institutionnels concernant le travail des femmes (roˆle de la se´gre´gation professionnelle des femmes, les ine´galite´s de salaires entre hommes et femmes, le leadership fe´minin en entreprise, les programmes publics d’accueil des enfants) influencent les perceptions individuelles de telles fac¸ons qu’elles augmentent le nombre de startups de femmes. En controˆlant les variations de structure d’opportunite´s qui existent entre les diffe´rents pays, nous montrons que les institutions (le leadership fe´minin en entreprise, les ine´galite´s de salaires entre hommes et femmes, et les de´penses publiques consacre´es a` l’accueil des enfants ) influencent indirectement – a` travers des perceptions concernant les sexes – la de´cision de cre´er une entreprise. European Journal of Development Research (2010) 22, 331–348. doi:10.1057/ejdr.2010.19 Keywords: female entrepreneurship; cultural institutions; cross-national research
Introduction Research indicates that women around the globe are starting new businesses on average at about two thirds the rate observed for men (Allen et al, 2008). Moreover, the size of the gender gap in business creation varies considerably across countries. Until recently, explanations for cross-national variations in the gender gap in business creation have focused primarily on structural factors such as unemployment, national wealth, economic growth and economic freedom (Verheul et al, 2006; Minniti and Nardone, 2007). Despite somewhat systematic variations of gender gaps in entrepreneurship with national wealth (that is, low/middle/high-income countries), considerable variation in the gender gap r 2010 European Association of Development Research and Training Institutes 0957-8811 European Journal of Development Research Vol. 22, 3, 331–348 www.palgrave-journals.com/ejdr/
Elam and Terjesen remains among countries with similar levels of national wealth. In fact, existing evidence shows that economic factors do not fully explain the observed gender patterns (Minniti et al, 2005; Baughn et al, 2006; Minniti and Nardone, 2007). Our study pursues this line of inquiry, investigating the importance of various types of gender-related cultural institutions on gender patterns of business start-up. Cultural beliefs, some scholars have argued, have emerged as a key distinguishing feature across advanced industrial nations (Esping-Andersen, 1999; Lenski, 2005). The structural foundations upon which the post-industrial nations are built have converged to include similar forms of productive technologies and economic structures that result in similar standards of living and lifestyles (Nolan and Lenski, 1999). Yet, important, measurable differences remain. For this reason, any attempt to understand cross-national variation must take into consideration the socio-cultural factors that distinguish one country from another. This is not an easy task, however, because cultural differences are not as simple as understanding how the individuals in one nation think differently compared to those in another. Culture is complex and enduring. What begins as an idea or a single belief quickly becomes institutionalized in terms of norms, ideals and expectations of appropriate behavior. Formal policies and belief systems are then born out of normative and idealized patterns of practice. Enduring institutions are created in the form of laws, regulations, and even physical buildings. This notion that institutions reflect points in a process of the institutionalization of ideas is a fundamental sociological paradigm. This theoretical perspective is not well established in entrepreneurship studies, in which institutional views typically fall into two distinct schools of thought – the very common institutional economic definition of ‘rules of the game’ (for example, North, 1990; Harper, 2003) or the more sociological definition of institutions as cultural products (for example, Powell and DiMaggio, 1991; Scott, 2003). In this article, we adopt an explicitly sociological definition of institutions as enduring cultural ideas external to the individual and progressively structured into accepted behavioral patterns and cultural practices, as well as into more codified and even material forms (Johnson, 2000). In this sense, institutions can vary from soft forms – such as values, beliefs, ideals and expectations – to harder, more material or structuring forms – including institutionalized norms and practices, such as labor force composition, industrial technologies, and government laws and policies. And finally, specific institutions, from informal to formal, can contradict one another. In the case of gender essentialist beliefs, the institutionalization of cultural beliefs works in basically the same way. What begins as an idea about how men and women behave or should behave becomes everyday practice by some majority of people, and over time becomes increasingly institutionalized in the rules and arrangements that define a given society. In this sense, many of the patterns and processes we observe in society are not only institutionalized, but also fundamentally gendered (that is, expressing or producing differences relating to gender). Importantly, the complexity and potential contradiction of informal and formal institutions may be observed in the ways in which gender essentialist beliefs are mitigated by the structural realities encountered by actors in everyday life. For example, in traditional societies in which gender essentialist beliefs often prevail, the requirements of economic survival and success may necessitate high levels of female labor force participation. Indeed, research findings on gender and work indicate that a range of institutional factors offer important explanations for gender differences in work patterns (Esping-Andersen, 1999; van der Lippe and van Dijk, 2002; Inglehart and Norris, 2003). At this point in time, 332
r 2010 European Association of Development Research and Training Institutes 0957-8811 European Journal of Development Research Vol. 22, 3, 331–348
Gendered Institutions and Business Creation however, very little is known about the relationship between national-level institutional arrangements and patterns of business creation for men and women (De Bruin et al, 2007). For example, Minniti and Nardone (2007) tested three socioeconomic factors: national wealth, economic growth rate and economic freedom, controlling for individual-level factors, reporting little explanation of start-up differences across gender. In another study, Baughn et al (2006) found that countries with higher levels of entrepreneurial activity also tend to have higher levels of female entrepreneurship. And finally, in a third study, Elam (2008) found that national belief in a traditional gender division of labor influences nascent entrepreneurship through key perceptions at the individual level. This article extends this research investigating the extent to which different types of gender-linked social/cultural institutions may mitigate the influence of gender essentialist beliefs on the decision to start a business for men and women. In this study, we draw on institutional and feminist perspectives of entrepreneurship to develop four hypotheses concerning the ways in which social institutions – from a specific belief about the gender division of labor to more concrete norms and institutional factors – influence gender patterns of business start-up. To test our hypotheses, we compiled data from several sources–individual data from the Global Entrepreneurship Monitor (GEM) and national-level data from the UN, World Bank, IMF, OECD and ISSP programs – which we analyzed using a series of two-level random intercept logistic regression models. Controls included individual measures of context, perceptions of self and environment, and national wealth, income inequality, and gender ideology. Below, we report results for cross-level interaction effects between the four test factors and the controls, and for postestimation probabilities. Finally, we conclude with implications for theory, practice and further research. This study makes several important contributions. First, it contributes to the research on the importance of those ‘enduring ideas’ embodied in cultural institutions for patterns of business start-up among women and men. Second, it answers calls for more crossnational comparative research on both gender and entrepreneurship (McManus, 2001). Comparative research holds a great deal of power for disentangling the complex mechanics of gender systems. And finally, it provides a model for testing the processes and mechanisms linking individual-level decisions and actions with macro-level factors. Multilevel modeling holds great promise for the investigation of the social and cultural processes driving the decision to start a business (Davidsson and Wiklund, 2001).
Theorizing Gender Patterns in Entrepreneurship Feminist analyses of gender patterns of work tend to focus on two key issues – occupational segregation and gender wage inequality. Writ large these terms refer to the basic institutional processes – the division of labor, the valuation of gender and the distribution of rewards. Women in most societies are responsible for housework and childcare – forms of work that are traditionally less valued than the tasks assigned to men (de Beauvoir, 1952; Williams and Best, 1990). In fact, historical evidence suggests that female gathering efforts in hunting and gathering cultures and, later, the sales of female domestic products, such as eggs, baked goods and other homemade products, often sustained families through good and bad times (Rosenfeld, 1985; Blumberg, 2004). Even today, research on the gender division of labor and the valuation of so-called ‘women’s work’ continues to result r 2010 European Association of Development Research and Training Institutes 0957-8811 European Journal of Development Research Vol. 22, 3, 331–348
333
Elam and Terjesen INDIVIDUAL LEVEL
COUNTRY LEVEL Gendered Institutions:
Direct
•Industry Employment •Female Leadership •Gender Wage Gap •Plublic Childcare
Indirect
Controlling for: National Wealth and Inequality National Gender Beliefs
Business Start
Social Capital Gender Perceptions
Controlling for: Economic and Cultural Capital
Figure 1: Model of hypotheses.
in lower rewards, if not in a total lack of recognition as ‘work’ at all (Entwisle et al, 1995; Charles and Grusky, 2004). Taking a cue from sociological theories of institutions, culture and the modern industrialized welfare state, we identify four types of institutions shaping women’s economic participation – female proportion of sectoral employment, gender wage inequality, presence of female business leadership and public expenditures on childcare – which influence gendered rates of business creation either directly or indirectly (see Figure 1). These factors further reflect aspects of both occupational segregation and the valuation of women’s economic contributions. Industrial Sectors The societal shifts in industrial sectors offer one way of looking at occupational segregation and its effects on gender. Sectoral shifts represent key transformations of productive technologies and explain a large portion of the variation in gender work patterns across countries (McManus, 2001; Charles and Grusky, 2004). Recent cross-national findings indicate that the size of the agricultural sector correlates positively with entrepreneurship rates for men (Reynolds et al, 2005), whereas the growth of the service sector may explain recent increases in female entrepreneurship growth rates (Terjesen, 2006). In addition, women’s labor is often rendered invisible in agricultural sectors. For example, evidence from women in development studies and farm women in the United States has shown that while ‘women’s work’ in these contexts tends to actually pay the bills, it is often not recognized as paid work (Rosenfeld, 1985; Entwisle et al, 1995). Women in the agricultural context may not feel especially confident or skilled at starting a business and/or may not recognize their efforts as business creation. Service work results in an entirely different work experience for women. These jobs are often stereotyped as more ‘feminine’ and well suited for female-type aptitudes and skill sets – especially work that involves emotional labor and caring (England et al, 2002). In service work, women may feel more confident about their abilities to start a business relative to more male-dominated or male-appropriate fields of work. Thus, we expect the following: Hypothesis 1a:
Countries with larger agricultural sectors are more likely to have higher rates of business start-up for men.
Hypothesis 1b:
Countries with larger service sectors are more likely to have higher rates of business start-up for women.
334
r 2010 European Association of Development Research and Training Institutes 0957-8811 European Journal of Development Research Vol. 22, 3, 331–348
Gendered Institutions and Business Creation Gender Wage Inequality Although occupational sex segregation describes the gender structure of the labor force, the gender wage gap captures the unequal rents generated in these activities. There is extensive variation across countries in the extent to which men and women are paid the same wage for the same job (Charles and Grusky, 2004; Charles and Bradley, 2009). Pay is, in effect, the primary outcome of the employment exchange relationship and women are generally disadvantaged partners. Wage restrictions, such as pay differentials, which affect one social group over another indicate fewer labor market options for the disadvantaged group. Across the world, restricted opportunities in the labor market have driven individuals to pursue self-employment (Dyer, 1994; McManus, 2000, 2001). This shift is especially prevalent among individuals earning the lowest income and living in the least-developed countries or those who have recently experienced economic transitions, such as by the fall of communism (for example, Acs et al, 2005). For example, evidence from the United States indicates that the class distribution of the self-employed varies significantly across gender (Devine, 1994; McManus, 2000, 2001). The distribution for men looks fairly normal, with most of the self-employed emerging out the middle class, whereas the distribution for women is concave, with more women starting businesses at the lowest and highest levels of income (Devine, 1994; McManus, 2000, 2001). Compared to men, women face greater barriers to employment in the formal labor market (ILO, 2004). Traditional expectations about women’s role in the household may discourage women from pursuing certain labor activities. Even when obtained, women’s formal jobs tend to be more temporary: women are often the first to be laid off and the last to be hired (ILO, 2004). Furthermore, women may seek flexible employment options that are more possible in self-employment than in the traditional wage/salary employment (Budig, 2006). Finally, women’s frustration with ‘glass ceilings’ and other forms of vertical segregation may motivate them to consider entrepreneurship (Leoni and Falk, 2010). Indeed, highly skilled women earn greater income in self-employment than their traditional wage and salary counterparts, whereas low-skilled women earn lower returns (Devine, 1994; Budig, 2006). On the basis of this reasoning, we expect: Hypothesis 2:
Countries with larger gender wage gaps will be more likely to have higher rates of business start-up for women.
Female Business Leadership A third component of economic opportunity is related to women’s business leadership. Although worldwide men and women are wage-employed in roughly equal numbers, women are significantly underrepresented in managerial roles, averaging only around 10 per cent (ILO, 2004). This discrepancy is even greater at the highest echelon, corporate boards, where women’s share averages 2 per cent, from 0.2 per cent in Japan to 22 per cent in Slovenia (Terjesen and Singh, 2009). The presence of female business leaders has a normative influence on society: legitimizing business as an acceptable, if not desirable, career option for women and increasing the value of women’s contributions in the marketplace. Indeed, a recent Catalyst study showed a significant positive correlation between female representation on corporate boards and the female representation in both line and staff positions 5 years later among r 2010 European Association of Development Research and Training Institutes 0957-8811 European Journal of Development Research Vol. 22, 3, 331–348
335
Elam and Terjesen Fortune 500 companies (Joy, 2008). Female business leadership may also change the way business is done – that is, it may change the cultural definitions of appropriate business practices and the ideal characteristics of business owners and managers. In this sense, higher rates of female business leadership may contribute to more women starting businesses, as the pursuit becomes a more obvious alternative and the marketplace becomes friendlier to women business owners. Hypothesis 3: Countries with higher numbers of female leaders are more likely to have higher rates of business start-up for women. Public Childcare Expenditures Government support for working mothers varies extensively across countries, even across industrialized countries (Chang, 2004). The purpose of these policies and programs is to enable women’s economic participation, regardless of ability to afford childcare services. Indeed, the gender work literature indicates that overall employment rates are higher in countries where state-based strategies (universal daycare, child tax subsidies) predominate and public-sector employment boosts female employment rates significantly (van der Lippe and van Dijk, 2002; Chang, 2004). However, in the case of business start-up, we encounter a dilemma. The countries with higher public expenditures on childcare also tend to be countries with large social welfare sectors and current research indicates that large welfare sectors are negatively correlated with entrepreneurial participation rates (Reynolds, Bygrave and Autio, 2004). Consequently, state-based childcare may discourage women from starting businesses because, although it successfully mediates the challenges to full-time employment faced by women in conservative and liberal economies, it also removes the ‘necessity’ motivation for women who seek flexible work alternatives. Indeed, there is an increasing trend in business creation among women in industrialized countries (Allen et al, 2008). For example, one US study found that women are starting part-time businesses at an increasing rate as an alternative to the inflexibility of full-time market work (Mainiero and Sullivan, 2006). Comparative studies on women’s work patterns suggest that as women are liberated from the necessity motivation, gender essentialism in the form of traditional gender beliefs or even idealistic expressions of gender may actually dissuade women from pursuing male-typed occupations, such as business start-up (Charles and Bradley, 2009). Consequently, we predict that: Hypothesis 4:
Countries with higher expenditures on public childcare provision have lower rates of start-up, especially for women.
Data and Methodology The individual-level data for the 11 countries (n ¼ 25 265) came from the 2001 GEM adult population survey data, which comprises nationally representative samples and responses to a single interview schedule agreed upon by all national research teams. The surveys were translated and collected by market survey firms with the appropriate experience and under the direct supervision and final review of the appointed national teams. (See Reynolds et al (2005) for an overview of GEM methodology.) Missing data and sampling issues required 336
r 2010 European Association of Development Research and Training Institutes 0957-8811 European Journal of Development Research Vol. 22, 3, 331–348
Gendered Institutions and Business Creation that the analysis be restricted to individuals in the age group 18-64 years in 11 countries. The countries in the sample included the United States, Germany, Denmark, Japan, Canada, Finland, Israel, Hungary, Poland, New Zealand and Portugal. Country-level data were collected from several sources of aggregate country-level data, including the United Nations 2005 Human Development Report, 2005 World Bank Indicators, 2002 International Social Survey Program (ISSP), 2004 United Nations Economic Commission for Europe (ECE), Cranfield Centre for Developing Women Business Leaders and the OECD Family Database. The measures selected, and discussed below in detail, were selected to most closely match the 2001 collection date of the individual-level data. Please note that the data sets listed include data collected in a range of years. When possible, 2000 or 2001 measures were collected. In the case of purely cultural measures, such as gender beliefs, precision was deemed less important as aggregate national cultural measures tend not to change much from year to year. Adopting a multi-level research model, we used two-level random coefficient logistic regression models to investigate the importance of gendered cultural institutions on nascent entrepreneurship rates for men and women in the multi-country sample, controlling for individual measures of context and perceptions of self and environment. Country-level test measures included sectoral employment, gender wage gap, female business leadership and public expenditures on childcare. We controlled for national wealth, income inequality (Gini), gender essentialist beliefs, and the individual-level interaction of gender and fear of business failure. Macro-micro interactions were further analyzed to explore possible interactions of institutional arrangements with gender and perceptions at the individual level. Dependent Variable Nascent entrepreneurship was our dependent variable. We defined nascent entrepreneurs as those individuals engaged in a business start-up, active over past 12 months and with expectations of full or part ownership. Thus, nascent entrepreneurship includes all individuals involved in a business start-up, active over past 12 months, with the expectation of full or part ownership. Descriptive analysis (Table 1) reveals that about 5 per cent of total respondents in this multi-country sample qualify as nascent entrepreneurs, with the proportion of nascent entrepreneurship ranging from 1 per cent in Japan to 14 per cent in New Zealand. Rates for women were consistently lower across all countries and significantly lower in most countries.
Test Variables We tested two industrial sector measures – proportion of the national labor force employed in agriculture and the proportion employed in services, both from 2004 United Nations ECE statistical database. The measure for gender wage inequality also came from the United Nations ECE database. We tested one measure of female business leadership – female representation among business leaders from the Cranfield Centre for Developing Women Business Leaders. This variable captures the percentage of the top 50 companies’ presidents and board members who are women. We tested one measure of public childcare expenditures from the OECD Family Database. This measure represents public expenditure on childcare as a percentage of 2003 GDP. See Table 1 for descriptive statistics by country. r 2010 European Association of Development Research and Training Institutes 0957-8811 European Journal of Development Research Vol. 22, 3, 331–348
337
338
r 2010 European Association of Development Research and Training Institutes 0957-8811 European Journal of Development Research Vol. 22, 3, 331–348
6.8 5.4 9.9 1.8 4.2 0.1 4.2 0.4
9.5 24.7 10.5 1.6 61.9 0.3 0.2 0.01
34.3 40.8 45.7 7.5 77.6 0.5 15.0 1.50 34.3 40.8 23.6 1.64 77.6 0.38 15.0 0.30
0.12 0.51 0.42 0.35 0.23 40.8 0.10 0.35 0.45 0.11 0.64 0.12 0.16 0.00 0.08 0.34 0.59 0.21 0.42 12.3 26.9 39.2 5.26 61.9 0.41 8.0 0.70
0.05 0.51 0.25 0.40 0.34 40.0 0.34 0.53 0.09 0.04 0.55 0.05 0.06 0.06 0.27 0.08 0.53 0.25 0.37 29.0 24.7 13.6 3.65 74.8 0.28 5.0 1.50
0.02 0.59 0.40 0.35 0.26 42.1 0.21 0.26 0.52 0.00 0.74 0.00 0.05 0.01 0.20 0.34 0.37 0.27 0.40 9.5 34.1 45.7 — — 0.38 8.0 0.01
0.04 0.50 0.25 0.46 0.29 38.8 0.22 0.67 0.11 0.00 0.52 0.03 0.12 0.07 0.26 0.11 0.34 0.46 0.36 25.4 28.3 20.5 2.34 71.0 0.48 6.0 0.90
0.05 0.53 0.34 0.39 0.27 42.3 0.40 0.38 0.22 0.00 0.71 0.00 0.29 0.00 0.00 0.24 0.33 0.47 0.40 19.2 36.2 19.8 7.52 69.7 0.31 14.0 0.20
0.14 0.52 0.24 0.53 0.23 40.9 0.35 0.33 0.30 0.03 0.63 0.17 0.04 0.07 0.08 0.42 0.64 0.30 0.52 25.1 24.9 30.9 4.52 67.5 0.54 0.2 0.40
0.01 0.51 0.22 0.42 0.36 41.1 0.07 0.49 0.43 0.01 0.53 0.03 0.10 0.25 0.09 0.05 0.10 0.18 0.16 27.1 33.1 10.5 2.03 70.6 0.37 13.0 0.30
0.07 0.50 0.21 0.52 0.27 39.3 0.14 0.31 0.50 0.05 0.58 0.17 0.04 0.07 0.14 0.28 0.51 0.28 0.34 18.2 38.5 30.6 — — 0.46 5.0 0.45
0.03 0.52 0.36 0.36 0.28 39.7 0.61 0.24 0.15 0.00 0.61 0.04 0.35 0.00 0.00 0.17 0.38 0.37 0.26
24.4 26.9 11.7 4.91 69.4 0.30 13.0 1.00
0.02 0.49 0.40 0.39 0.20 41.6 0.21 0.23 0.57 0.00 0.68 0.00 0.06 0.12 0.11 0.40 0.37 0.35 0.49
19.8 35.5 16.9 — — 0.47 10.0 0.09
0.02 0.55 0.30 0.37 0.33 36.1 0.07 0.41 0.52 0.00 0.63 0.00 0.08 0.13 0.16 0.15 0.31 0.28 0.30
US HU DK PL DE NZ JP CA PT FI IL Mean Mean Mean Mean Mean Mean Mean Mean Mean Mean Mean
Note: Bold and italic figures indicate lowest and highest, respectively, across countries in sample.
22.9 31.6 23.9 3.6 70.6 0.4 8.5 0.6
23253 23253 23253 17832 17832 23253 23253 23253
1 1 1 1 1 64 1 1 1 1 1 1 1 1 1 1 1 1 1
Min Max
Country National wealth (per capita gdp) National income inequality (gini) National gender ideology % Labor force in agriculture % Labor force in services Gender pay gap (1)female/male) Female business leadership % Public expenditures on childcare
sd
23253 0.05 0.22 0 23253 0.52 0.50 0 17905 0.32 0.47 0 17905 0.41 0.49 0 17905 0.28 0.45 0 23253 40.54 12.97 18 22087 0.27 0.44 0 22087 0.38 0.49 0 22087 0.33 0.47 0 22087 0.02 0.15 0 23253 0.63 0.48 0 23057 0.05 0.21 0 23057 0.15 0.36 0 23057 0.06 0.23 0 23057 0.10 0.31 0 23073 0.23 0.42 0 23217 0.40 0.49 0 23222 0.33 0.47 0 23248 0.37 0.48 0
Mean
Individual Active start-up (DV) Female Top third income Mid third income Low third income Respondent’s age Less than secondary educ Secondary educ degree Some post-secondary educ College graduate Employed full time Employed part time Unemployed Homemaker Retired or student Expects to see good opportunities Has skills to start a business Fears new business failure Knows other entrepreneur(s)
n
Table 1: Descriptives for GEM 2001 sample of 11 countries (bold=min; italic=max)
Elam and Terjesen
Gendered Institutions and Business Creation
Control Variables In each country, individuals are distributed across a measurable opportunity structure. In this article, we define opportunity structure according to Bourdieu’s (1986) capital framework, which defines individual social positions in terms of four forms of capital – economic, social, cultural and symbolic. We included two binary measures of economic capital, derived from a ranked categorical variable of household income recoded into relative thirds within a given country. The measures included in the regression models are in the lower third and in the upper third of the national income distribution. In anticipation of a possible curvilinear relationship between household income and nascent entrepreneurship, we used middle income as the reference category. We also controlled for two types of cultural capital measures: institutionalized (education, experience and work status) and habitus (perceptions). A continuous measure for education was not available, and therefore the level of education is tested directly with a set of three dichotomous measures – secondary education or less, post-secondary education and graduate education. Following conventional practice, we use age as a proxy for work experience. Age, of course, is not necessarily the best way to measure work experience, especially in cases in which women take time out from paid work to care for children or the home. For this reason, a measure of work status is also included in the form of four dichotomous variables: part time, homemaker, retired/student and unemployed, with full-time employment as the reference category. Three dichotomous measures of habitus were controlled, including perceptions of good start-up opportunities, perception of start-up skills and fear of failure. In addition, we tested one dichotomous measure of social capital – personally knowing an entrepreneur over the past 12 months. As discussed earlier, women experience significant normative disadvantages in the labor market. As such, we frame gender status as a possible source of legitimacy. We use a dichotomous measure of gender where 1 ¼ female and 0 ¼ male. With regard to the macro-level controls, national wealth was measured by 2001 per capita GDP (GDPPC), a standardized national statistic drawn from the United Nations 2005 Human Development Report. We used 2001 Gini index scores from the 2005 World Bank Indicators to measure income inequality. Gini scores measure the extent to which the distribution of income among individuals or households within a country deviates from a perfectly equitable distribution on a scale of 0 (perfectly equal) to 100 (completely unequal). The final country-level measure, gender culture, was belief in a traditional gender division of labor. The national scores calculated were based on responses to one attitudinal question in the ISSP 2002 Family and Changing Gender Roles module – ‘A man’s job is to earn money, a women’s job is to look after the home and family.’
Results Results of the analyses suggest that institutional measures impact nascent entrepreneurship rates differently for men and women. The results for a set of 6 two-level random coefficient logistic regression models are presented in Table 2. In the interests of establishing a baseline for comparison across models, we first review the results r 2010 European Association of Development Research and Training Institutes 0957-8811 European Journal of Development Research Vol. 22, 3, 331–348
339
Elam and Terjesen from Models 1 and 2 before proceeding to a presentation of the results for our hypotheses tests. Model 1 is the basic reference model and shows the extent of the country variation (0.440) before the controls and additional variables are added. Model 2 includes all of the controls. We controlled for social position because we wanted to compare individuals in similar structural locations across countries. We also controlled for three macro-level measures – national wealth, national income inequality and national gender belief in traditional gender roles. The controls explain about 50 per cent of the cross-national variation in nascent entrepreneurship and 18 per cent of the gender variation. Below, we present our results according to each of the four hypotheses. Industrial Sectors In H1a, we hypothesized that the share of employment in agriculture was an important predictor of business start-up for men. In H1b, we hypothesized that the importance of the share of employment in services was an important predictor of business start-up for women. Women work hard in agriculture sectors, but their work is often counted as family labor and is rendered invisible in formal employment counts, despite a possible entrepreneurial character. Manufacturing work is typically characterized as work more suitable for men. In contrast to both agricultural work and manufacturing work, services work is characterized as suitable for women. As shown in Tables 2 and 3, the results offer limited support for H1a and H1b. Model 3 in Table 2 shows no direct effect of agricultural sector or services sector employment on business start-up and no indirect effects (or interaction effects) for service sector. There is, however, one significant interaction effect for perception of start-up skills and proportion of employment in agriculture. This finding indicates that, among respondents who claimed the skills to start a business, the odds of being a nascent entrepreneur decreases by about 13 per cent for each additional percentage point increase in the share of national employment in agriculture. This result stands in contrast to previous findings on the relationship between the extent of agricultural employment and entrepreneurship participation rates. One likely explanation can be found in the controls. Much of the crossnational variation in nascent entrepreneurship is controlled for by social position and national structural characteristics. Holding these variables constant, it is very possible that general correlations between variables may shift and weaken. Further, as indicated by the results in Tables 2 and 3, controlling for the employment share in agriculture and services sectors, with manufacturing as the reference category, produced models that explain 80–90 per cent of the cross-national variation in business start-up and over 90 per cent of the gender variation in business start-up. The results also show that variations in industrial sector employment shares likely play a role in the relationship between work status patterns and the likelihood of business start-up. Such findings suggest that share of employment across industrial sectors is an important source of gendered patterns of start-up. The importance of this finding is further seen in the lack of a significant direct gender effect in either of the relevant interaction models. In fact, analyses of the Empirical Bayes post-estimation probabilities (results not shown) indicate that these sectoral employment measures are important explanations of variation in entrepreneurship across countries. Controlling for these measures appears to change predicted probabilities of business start-up for men and women in opposite 340
r 2010 European Association of Development Research and Training Institutes 0957-8811 European Journal of Development Research Vol. 22, 3, 331–348
r 2010 European Association of Development Research and Training Institutes 0957-8811 European Journal of Development Research Vol. 22, 3, 331–348
4871.07 25 265 11
log likelihood N (i) N (j)
Note: *Po0.05, **Po0.01, ***Po0.001.
0.440 0.051
— — — — — — — —
National wealth (per capita gdp) National income inequality (gini) National gender ideology % Labor force in agriculture % Labor force in services Gender pay gap (1female/male) Female business leadership % Public expenditures on childcare
Var (country) Var (gender)
0.47 — — — — — — — — — — — — — — — — —
exp(b)
se
— — —
0.200 0.041
— — — — — — — —
0.049*** — — — — — — — — — — — — — — — — —
Model 1
Female Top third income Mid third income Low third income Respondent’s age Less than secondary educ Secondary educ degree Some post-secondary educ College graduate Employed full time Employed part time Unemployed Homemaker Retired or student Expects to see good opportunities Has skills to start a business Fears new business failure Knows other entrepreneur(s)
Active business start
2963.33 18 488 11
0.220 0.011
1.00 1.01 1.02 — — — — —
0.78 1.16 1.06 0.99 1.02 1.07 1.24 1.42 0.80 0.74 0.53 2.91 5.77 0.59 2.63 1.00 1.01 1.02
exp(b)
se
— — —
0.104 0.024
0.020 0.025 0.013 — — — — —
0.079** 0.112 0.088 0.003*** 0.106 0.115 0.244 0.186** 0.101 0.189 0.091*** 0.214*** 0.584*** 0.053*** 0.212*** 0.020 0.025 0.013
Model 2
2346.16 13 038 8
0.073 0.019
0.90 1.11 1.01 0.85 1.06 — — —
0.89 1.18 0.99 0.99 1.05 1.11 1.27 1.23 0.75 0.45 0.54 2.94 5.45 0.65 2.58 0.90 1.11 1.01
exp(b)
se
— — —
0.067 0.028
0.063 0.024*** 0.014 0.075 0.085 — — —
0.092 0.128 0.092 0.003*** 0.121 0.130 0.253 0.179 0.108* 0.153* 0.108** 0.241*** 0.621*** 0.064*** 0.228*** 0.063 0.024*** 0.014
Model 3
Table 2: Odds ratios for 2-level random regression models on nascent entrepreneurship
2738.74 16 701 11
0.218 0.032
1.01 1.02 1.02 — — 1.29 — —
0.79 1.21 1.05 0.99 0.99 1.06 1.22 1.25 0.75 0.68 0.54 2.94 5.44 0.62 2.61 1.01 1.02 1.02
exp(b)
se
— — —
0.125 0.050
0.023 0.029 0.020 — — 2.799 — —
0.087* 0.123* 0.091 0.003*** 0.107 0.118 0.241 0.178 0.100* 0.184 0.096*** 0.226*** 0.571*** 0.056*** 0.216*** 0.023 0.029 0.020
Model 4
2738.26 16 701 11
0.212 0.005
1.04 1.01 1.05 — — — 1.08 —
0.88 1.22 1.06 0.99 0.99 1.05 1.21 1.27 0.76 0.69 0.53 2.94 5.43 0.62 2.61 1.04 1.01 1.05
exp(b)
se
— — —
0.128 0.019
0.023 0.037 0.019** — — — 0.050 —
0.108 0.123* 0.092 0.003*** 0.106 0.117 0.239 0.181 0.101* 0.185 0.095*** 0.225*** 0.570*** 0.056*** 0.216*** 0.023 0.037 0.019**
Model 5
2738.08 16 701 11
0.207 0.044
1.02 0.98 1.02 — — — — 0.58
0.78 1.21 1.06 0.99 0.99 1.05 1.21 1.24 0.75 0.68 0.54 2.94 5.45 0.62 2.61 1.02 0.98 1.02
exp(b)
se
— — —
0.123 0.065
0.025 0.042 0.017 — — — — 0.328
0.087* 0.123 0.091 0.003*** 0.106 0.117 0.239 0.177 0.100* 0.183 0.096*** 0.226*** 0.573*** 0.056*** 0.217*** 0.025 0.042 0.017
Model 6
Gendered Institutions and Business Creation
341
Elam and Terjesen directions. There is a clear interaction in some countries – but there are three distinct patterns. In the United States, Finland and Denmark, controlling for labor sectors increases probabilities of business start-up for women and decreases them for men, increasing the gender gap. The opposite is true in New Zealand and Canada where controls for industry decreases gender differences and, finally, the probabilities change in similar directions for men and women in Japan and Poland. To best understand these changes in predicted probability, differences in industry sector share of overall employment as well as occupational sex segregation must be examined. Gender Wage Inequality In H2, we predicted that countries with larger gender wage gaps would see higher rates of nascent entrepreneurship for women. The results in Tables 2 and 3 show do not support this hypothesis. No direct effects appear in Model 4 and no indirect effects appear in the pay gap model in Table 3. This result was surprising as a direct effect did emerge in our preliminary analysis of these data and has appeared in analyses of later waves of GEM data. Evidence of multicollinearity is indicated by inflated standard errors. For these reasons, we believe that the results may reflect data quality issues with the macro measures and that further inquiry is warranted on this point – with particular attention to later waves of GEM data. As these findings stand, the results indicate that controlling for the level of gender wage inequality may actually decrease the amount of gender variation explained by the reference models. Countries with similar levels of gender wage inequality may see quite different levels of female versus male business start-up, suggesting exogenous factors at work. For example, the United States and Canada share similar levels of gender wage inequality, but the gender gap in business start-up is significantly larger in Canada than in the United States, which is likely because of factors not included in the test models. Female Business Leadership In H3, we predicted that countries with higher proportions of female business leaders would see higher rates of business start-up for women. The results of our analysis offer strong support for this hypothesis. Results for Model 5, shown in Table 2, reveal no significant direct effect for female business leadership. However, the results in Table 3 show interaction effects for female business leadership with gender and perception of good business opportunities. Also, although female business leadership and related interactions do not appear to explain much in the way of country variation in nascent entrepreneurship, the models in question do explain about 90 per cent of the gender variation; in other words, an additional 12 per cent beyond the 78 per cent of gender variation explained by individual controls alone. The significance of this finding is also found in that gender no longer has a significant direct effect in these models. Public Expenditures on Childcare In H4, we predicted that public expenditure on childcare would decrease (female) business start-up in one of two ways. Higher public expenditures on childcare may free parents, especially mothers, to start businesses. However, higher public expenditures on childcare may also serve as a proxy for larger welfare state structures and institutions, which are 342
r 2010 European Association of Development Research and Training Institutes 0957-8811 European Journal of Development Research Vol. 22, 3, 331–348
Notes: Results for controls not shown; *Po0.05.
— — —
2339.35 13 038 8
log likelihood N (i) N (j)
0.050 0.043 0.053* 0.054 0.052 0.051 0.013
1.030 0.938 0.874 1.020 1.030
SE
0.050 0.004
female expects to see good opportunities has skills to start a business fears new business failure knows other entrepreneur(s)
exp(b)
agric sector
Var (country) Var (gender)
X X X X X
Active business start
2339.35 13 038 8
0.050 0.004
1.033 0.972 0.976 1.047 0.993
exp(b)
SE
— — —
0.051 0.013
0.025 0.022 0.033 0.031 0.023
svc sector
2736.58 16 701 11
0.231 0.038
0.476 0.842 13.015 2.338 0.847
exp(b)
SE
— — —
0.128 0.056
0.708 1.008 19.393 3.248 1.107
paygap
2731.32 16 701 11
0.192 0.000
1.046 1.045 0.954 1.022 1.003
exp(b)
fbus
Table 3: Interaction effects only in odds ratios for 2-level random regression models on nascent entrepreneurship
— — —
0.091 0.004
0.020* 0.020* 0.024 0.024 0.020
SE
2732.24 16 701 11
0.188 0.023
0.687 0.618 1.619 0.794 0.730
exp(b)
child
— — —
0.100 0.043
0.173 0.134* 0.471 0.207 0.172
SE
Gendered Institutions and Business Creation
r 2010 European Association of Development Research and Training Institutes 0957-8811 European Journal of Development Research Vol. 22, 3, 331–348
343
Elam and Terjesen typically correlated with lower rates of entrepreneurship for both women and men. The results shown in Tables 2 and 3 offer some support for H4. Although we found no direct effect of GDP expenditures on public childcare (Model 6), we did find an indirect effect for public childcare expenditures through perception of good business opportunities. The odds of business start-up were significantly decreased by 38 per cent per percent GDP expenditure on childcare for those who expected to see a business opportunity within the near future, suggesting that welfare provisions may moderate the association between perception of opportunity and business start-up for both genders. Across all the countries in the sample, men were significantly more likely than women to expect good business opportunities in the near future; in this way, public expenditures on childcare work through perceptions to influence gendered patterns of business start-up. Post-estimation analysis of Empirical Bayes probabilities offers further evidence for the complexity of these processes. Male and female probabilities are similarly affected by the addition of childcare expenditure to the regression analyses in New Zealand. In other countries such as Japan, Canada and the United States, male and female probabilities of nascent entrepreneurship change in opposite directions. In the remaining majority of countries in the sample, the addition of public childcare expenditure decreased the gender gap in the predicted probability of business start-up. HLM modeling across countries is a complex task. Hence, limitations in both the data and the analysis are unavoidable. The measures used in this study reflected the compromise between the available data, the measure/control of key variables from previous findings, and, of course, the theoretical model of business start-up presented. The macrolevel variables all suffer from a number of missing data issues, which drastically reduced the number of countries we were able to use in our analyses. A greater number of countries would have produced more reliable random variance estimates and a greater diversity of countries would have offered a richer set of data in terms of the theoretical model. As it stands, the countries included in the analysis were quite well developed. Poland had the lowest level of national wealth at this point in time and had the highest national score in gender essentialism (that is, the highest percentage of individuals expressing a belief in a traditional gender division of labor). Still, the countries included offered enough variation to establish a baseline for this kind of modeling in investigations of differential start-up rates across countries. In addition, there are some variables that are particularly critical to the study of gender and entrepreneurship, which were not available in the individual-level data and deserve mention here. Industry, for one, is a strong predictor of business start-up and performance. Family context (for example, marriage and presence of children) are also important examples and would have improved the quality of the study. Additional social capital measures, such as support of friends and family and the availability of expert advisors (beyond other entrepreneurs), are important factors for the decision to start a business. And, finally, fewer missing country-level data would have improved the quality of this study by ensuring a more consistent country sample possible across all models.
Discussion Perhaps the clearest general evidence offered by these results is that the effects of various gender institutions on rates of business start-up for men and women are highly dependent on the larger institutional context. No direct effects were found for any of the test 344
r 2010 European Association of Development Research and Training Institutes 0957-8811 European Journal of Development Research Vol. 22, 3, 331–348
Gendered Institutions and Business Creation variables. However, indirect effects were found for industrial sector employment share, female business leadership and public expenditures on childcare. These findings suggest that gender institutions work through factors, such as key perceptions/expectations and gender status at the individual level to influence patterns of business creation. For example, sectoral employment factors and female business leadership appear to explain the direct effect of gender on rates of business start-up. Sectoral employment patterns, in particular, appear to be important explanations of variations in nascent entrepreneurship for both gender and countries, but especially for variation across countries. Analyses of post-estimation probabilities further confirm this finding, reflecting a fair amount of consistency in probabilities of business start-up for all countries across all models. In other words, the countries tended to hold their rank in rates of entrepreneurship across all the models. The biggest changes, in fact, occurred between the first two models in Table 2 – with the addition of individual controls for the distribution of individuals across the national opportunity structure. Indeed, the results offer considerable evidence in favor of the importance of variations in opportunity structures and individual social positions as an explanation of cross-national patterns of entrepreneurship rates. The individual-level factors alone explained about 40 per cent of the variation in business start-up across the 11 countries in the sample. In a larger sample with a broader range of countries, we would expect to see an even greater proportion of the variation across countries explained by variations in opportunity structure. In terms of explaining gender differences in business start-up, the results illustrate the importance of gender differences in both opportunity structure and individual perceptions for explaining the lower overall participation rates for women. Previous research has shown that women face considerable disadvantage in agricultural contexts in which their work tends to be invisible, as in the case of farming families, or, unlikely, as in the cases of manual labor requiring great physical strength or capital intensive farm equipment. This situation, of course, relates to the perceptions that women experience about themselves and their environments. In this sense, service work should be advantageous to women, yet we see no positive effect here, only an absence of negative effects. An important point to note is that, although these industry employment share factors may not interact directly with gender, they may affect variables such as perceptions, social capital and other individual factors that help to explain much of the observed gender differences in entrepreneurship participation rates. Analysis of post-estimation probabilities suggests that the test variables are differentially related to male and female probabilities of nascent entrepreneurship. Overall, the decision to actively start a business appears more sensitive to the test variables in United States ( þ ), Germany (), New Zealand () and Canada ( þ ), compared to the other countries in the sample. Female business start-up appears more sensitive to test variables in United States ( þ ), Germany (), New Zealand (), Finland ( þ ) and Canada ( þ ), compared to the other countries in the sample. Finally, male business start-up appears more sensitive to test variables in Germany (), New Zealand (), Finland () and Canada ( þ ), compared to the other countries in the sample. Conceptually, these patterns appear to be expressions of three key propositions: (1) individuals in some countries are more influenced by institutional factors than those from other countries; (2) men and women may respond differently to institutional factors; and (3) the extent to which male and female entrepreneurs respond differently to institutional environment may depend on aspects of national culture and institutions unaccounted for. r 2010 European Association of Development Research and Training Institutes 0957-8811 European Journal of Development Research Vol. 22, 3, 331–348
345
Elam and Terjesen
Conclusion We began this article with the argument that culture is increasingly institutionalized into social structures or institutions of varying degrees that mediate the relationship between culture at the national level of analysis and behavior at the individual level of analysis. We also noted that cultural institutions ranging from ideas and beliefs to more structuring forms, such as established practices and government benefits, can contradict one another in terms of the effect on individual-level behavior. For gender differences, we argued, the key institutions likely to mediate the relationship between traditional gender beliefs and business start-up include industry sector employment share, gender wage gap, female business leadership and public expenditures on childcare. The findings suggest that gendered cultural institutions from ideology to more tangible patterns and practices offer important explanations for business creation. Importantly, the industry composition of a given economy and the gender division of labor associated with jobs in each industry sector may work through individual-level factors, such as the resources available for business start-up, including gender status as a source of legitimacy, or key perceptions tied to the propensity to start a business. The findings in this study also suggest that factors such as female business leadership and the availability of statesupported childcare may also work through individual-level factors to influence the rate of business start-up. In this study, we used an ambitious form of multi-level modeling to explore the question of how gendered institutions may moderate the relationship between gender and the decision to start a business at the individual-level of analysis for men and women. However, the methods are somewhat new to this area of inquiry (in entrepreneurship studies, if not sociology) and require further application. Although we have the methods, we still lack the theories to support the use of these methods in cross-national studies of entrepreneurship. Certainly, multi-level theories and research models are needed to explore linkages within and across levels of analysis. In particular, more attention needs to be paid to the definitions of culture and institutions, to overcome the confusion often created in interdisciplinary fields with parallel conceptualizations and co-opted terminology. And finally, there is a strong need to collect better comparative data, at both the micro and macro levels of analysis. Better data in the form of broader, more diverse, samples of countries will provide better fodder for analysis and better support for the efforts of sociologists and management scholars to continue exploring the ‘gray matter’ between polarized dichotomies, such as agency and structure, culture and structure, and so on. We believe that better theory will lead to better data. For example, good data sets exist for the measurement of informal institutions (for example, attitudes, beliefs and values) but not for more formal institutional arrangements, particularly with variables linked to gender. As multi-level theories of gender and institutional arrangements improve, and better data become available, there is a greater understanding of how status characteristics such as gender work to produce differential patterns of behavior across society and the economy. In and of itself, gender is a complex institution in society. Although many of us understand that female role models, and the appropriate compensation levels, are important for incentivizing women to achieve economically, factoring in how these variables interact with industry/occupational employment patterns and with policy initiatives to create specific outcomes is much more challenging. The development of better theories, methods and data will help us achieve more refined understandings of how gendered institutions serve as mechanisms linking the individual gendered experience to the larger cultural landscape. 346
r 2010 European Association of Development Research and Training Institutes 0957-8811 European Journal of Development Research Vol. 22, 3, 331–348
Gendered Institutions and Business Creation
Acknowledgement An earlier version of this article appears in the BCERC 2007 Foundations of Entrepreneurship Research proceedings. We thank Per Davidsson, Mike Peng, Roxanne Zolin and participants at the 2007 BCERC, 2007 Global Entrepreneurship Monitor and 2008 UNU-WIDER conferences for helpful feedback.
References Acs, Z.J., Arenius, P., Hay, M. and Minniti, M. (2005) Global entrepreneurship monitor 2004 executive report. In: M. Hancock (ed.) Global Entrepreneurship Monitor. Babson Park, MA: Babson College/London, UK: London Business School. Allen, I.E., Elam, A., Langowitz, N. and Dean, M. (2008) GEM 2007 Report on Women and Entrepreneurship Global Entrepreneurship Monitor Program. Babson College. Baughn, C.C., Chua, B.-L. and Neupert, K.E. (2006) The normative context for women’s participation in entrepreneurship: A multicountry study. Entrepreneurship: Theory & Practice 30(5): 687–708. Blumberg, R.L. (2004) Extending Lenski’s schema to hold up both halves of the sky: A theory guided way of conceptualizing agrarian societies that illuminates a puzzle about gender stratification. Sociological Theory 22(2): 278–291. Bourdieu, P. (1986) The forms of capital. In: J. Richardson (ed.) Handbook of Theory and Practice in the Sociology of Education. Westport, CT: Greenwood Press. Budig, M. (2006) Intersections on the road to self-employment: Gender, family, and occupational class. Social Forces 84(4): 2223–2239. Chang, M.L. (2004) Growing pains: Cross-national variation in sex segregation in sixteen developing countries. American Sociological Review 69(1): 114–137. Charles, M. and Bradley, K. (2009) Indulging our gendered selves? Sex segregation by field of study in 44 countries. American Journal of Sociology 114(4): 924–976. Charles, M. and Grusky, D.B. (2004) Occupational Ghettos: the Worldwide Segregation of Women and Men. Stanford, Ca: Stanford University Press. Davidsson, P. and Wiklund, J. (2001) Levels of analysis in entrepreneurship research: Current research practice and suggestions for the future. Entrepreneurship: Theory & Practice 25(4): 81. De Beauvoir, S. 1953 (1952) The Second Sex. New York: Alfred A. Knopf. De Bruin, A., Brush, C. and Welter, F. (2007) Advancing a framework for coherent research on women’s entrepreneurship. Entrepreneurship Theory and Practice 31(3): 323–339. Devine, T.J. (1994) Changes in wage-and-salary returns to skill and the recent rise in female selfemployment. American Economic Review 84: 108. Dyer, J.W.G. (1994) Toward a theory of entrepreneurial careers. Entrepreneurship: Theory & Practice 19: 7–21. Elam, A.B. (2008) Gender and Entrepreneurship: A Multilevel Model & Analysis. Cheltenham, UK; Northampton, MA: Edward Elgar. England, P., Budig, M. and Folbre, N. (2002) Wages of virtue: The relative pay of care work. Social Problems 49(4): 455–473. Entwisle, B., Henderson, G.E., Short, S.E., Bouma, J. and Zhai, F. (1995) Gender and family businesses in rural China. American Sociological Review 60(1): 36. Esping-Andersen, G. (1999) Social Foundations of Postindustrial Economies. Oxford, UK; New York: Oxford University Press. Harper, D.A. (2003) Foundations of Entrepreneurship and Economic Development. London: Routledge. Inglehart, R. and Norris, P. (2003) Rising Tide: Gender Equality and Cultural Change Around the World. Cambridge, UK; New York: Cambridge University Press. International Labor Organization (ILO). (2004) Global Employment Trends for Women. Geneva: ILO. Johnson, A.G. (2000) The Blackwell Dictionary of Sociology: A User’s Guide to Sociological Language. Malden, Ma: Blackwell Publishers. r 2010 European Association of Development Research and Training Institutes 0957-8811 European Journal of Development Research Vol. 22, 3, 331–348
347
Elam and Terjesen Joy, L. (2008) Advancing Women Leaders: The Connection between Women Board Directors and Women Corporate Officers. Catalyst Report. Lenski, G.E. (2005) Ecological-Evolutionary Theory: Principles and Applications. Boulder, CO: Paradigm Publishers. Leoni, T. and Falk, M. (2010) Gender and field of study as determinants of self-employment. Small Business Economics 34(2): 167–185. Mainiero, L. and Sullivan, S. (2006) The Opt Out Revolt: Why People are Leaving Companies to Create Kaleidoscope Careers. Mountain View, CA: Davies-Black. McManus, P.A. (2000) Market, state, and the quality of new self-employment jobs among men in the U.S. and Western Germany. Social Forces 783 (March): 865–905. McManus, P.A. (2001) Women’s participation in self-employment in Western industrialized nations. International Journal of Sociology 31(2): 70–98. Minniti, M., Arenius, P. and Langowitz, N. (2005) Global Entrepreneurship Monitor 2004 Report on Women and Entrepreneurship. Babson College and London Business School. Minniti, M. and Nardone, C. (2007) Being in someone else’s shoes: The role of gender in nascent entrepreneurship. Small Business Economics 28(2/3): 223–238. Nolan, P. and Lenski, G.E. (1999) Human Societies: An Introduction to Macrosociology, 8th edn. New York: McGraw-Hill College. North, D.C. (1990) Institutions, Institutional Change, and Economic Performance. Cambridge, UK; New York: Cambridge University Press. Powell, W.W. and DiMaggio, P. (1991) The New Institutionalism in Organizational Analysis. Chicago, IL: University of Chicago Press. Reynolds, P., Bosma, N., Autio, E., Hunt, S., Bono, N.D. and Servais, I. (2005) Global entrepreneurship monitor: Data collection design and implementation 1998–2003. Small Business Economics 24(3): 205–231. Reynolds, P.D., Bygrave, W. and Autio, E. (2004) GEM 2003 Executive Report. Babson, MA; Kansas City, MO: Babson College; Kauffman Foundation. Rosenfeld, R. (1985) Farm Women: Work, Farm, and Family in the United States. Chapel Hill, NC: University of North Carolina Press. Scott, W.R. (2003) Organizations: Rational, Natural, and Open Systems, 5th edn. Upper Saddle River, NJ: Prentice-Hall. Terjesen, S. (2006) Entrepreneurs’ transitions from corporate life to own ventures: leveraging human capital & social capital to establish new businesses. Thesis, Cranfield University. Terjesen, S. and Singh, V. (2009) Female presence on corporate boards: A multi-country study of environmental context. Journal of Business Ethics 83(1): 55–63. van der Lippe, T. and van Dijk, L. (2002) Comparative research on women’s employment. Annual Review of Sociology 28: 221–241. Verheul, I., Van Stel, A. and Thurik, R. (2006) Explaining female and male entrepreneurship at the country level. Entrepreneurship & Regional Development 18(2): 151–183. Williams, J.E. and Best, D.L. (1990) Measuring Sex Stereotypes: A Thirty Nation Study. Beverly Hills, CA: Sage.
348
r 2010 European Association of Development Research and Training Institutes 0957-8811 European Journal of Development Research Vol. 22, 3, 331–348
Copyright of European Journal of Development Research is the property of Palgrave Macmillan Ltd. and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.