شرح تنظیمات و مشارکت واقعی در خود اشتغالی: جنسیت و شخصیت کارآفرینی

This paper investigates an essential aspect of the entrepreneurial personality: why women’s self-employment rates are consistently lower than those of men. It has three focal points. It discriminates between the preference for self-employment and actual involvement in self-employment using a two (probit) equation model. It makes a systematic distinction between different ways in which gender influences the preference for and actual involvement in self-employment (mediation and moderation). It includes perceived ability as a potential driver of self-employment next to risk attitude, self-employed parents and other socio-demographic drivers. A representative data set of more than 8000 individuals from 29 countries (25 EU Member States, US, Norway, Iceland and Liechtenstein) is used (the 2004 Flash Eurobarometer survey). The findings show that women’s lower preference for becoming self-employed plays an important role in explaining their lower involvement in self-employment and that a gender effect remains that may point at gender-based obstacles to entrepreneurship.

Now that it is widely established that entrepreneurship is important for improving economic growth,1 policy-makers have been searching for ways to encourage groups of individuals that are underrepresented in the entrepreneurial population to start-up businesses (European Commission, 2002). In this quest for the entrepreneurial personality gender issues play a central role. Women are seen as an important potential resource for communities and regions aiming to expand their economies. Globally, women are less likely than men to behave entrepreneurially, whether this is measured in terms of newly founded or established businesses (Minniti et al., 2005 and Reynolds et al., 2002). Separating the different stages of entrepreneurship, such as the cognitive and behavioral stages, enables us to gain insight into the question of why some people become entrepreneurs and others do not (Baron, 2004). The decision to become an entrepreneur is traditionally seen as an occupational decision with two outcomes: to engage in entrepreneurial activity or to refrain from it. This ‘static’ perspective is challenged by a ‘dynamic’ approach that views entrepreneurship as a process consisting of several stages (Reynolds, 1997). For example, one can discriminate between pre-birth, birth, and post-natal stages of formation, where pre-birth is often referred to as latent or nascent entrepreneurship (Blanchflower, Oswald, & Stutzer, 2001; Masuda, 2006 and Gelderen et al., 2005).2 These stages of entrepreneurship may again have different antecedents (Davidsson and Honig, 2003, Grilo and Thurik, 2005b and Grilo and Thurik, 2008; Van der Zwan, Thurik, & Grilo, 2010). To understand why women are less likely to engage in entrepreneurial activity, one should investigate how they perform at these different stages of the entrepreneurial process and determine at what stage women start to lag behind and why. Not only are women less likely to become involved in entrepreneurship, but they also appear less interested in entrepreneurship (Blanchflower et al., 2001, Grilo and Irigoyen, 2006, Grilo and Thurik, 2005a and Grilo and Thurik, 2008). This lower preference for entrepreneurship among women may partially explain their lower level of entrepreneurial activity, suggesting that fostering female entrepreneurship should focus not just on the action stages of entrepreneurship but also on earlier attitudinal and decision stages. Hence, to establish to what extent women’s relatively low level of participation in entrepreneurial activity is driven by their lower preferences for entrepreneurship or, alternatively, by other factors (such as those related to a lower ability to become an entrepreneur), we investigate the antecedents of entrepreneurial preferences and entrepreneurial activity for both women and men. We link latent to actual entrepreneurial activity and examine how gender influences the relationship between these two stages. The entrepreneurial process is treated as a two-step procedure: the cognitive stage of ‘wanting it’ and the behavioral stage of ‘doing it’. Most studies investigating gender effects on entrepreneurship include gender as a dummy variable. Although this approach has its merits, it does not provide us with information on the origin of gender differences. By distinguishing between mediation and moderation effects3 on the decision and action stages of entrepreneurship, we aim to find out whether the lower female entrepreneurial activity rate can be attributed to a lower preference of women for becoming entrepreneurs or, alternatively, to the existence of gender differences with respect to other (ability) factors that influence engagement in entrepreneurial activity. In sum, the contribution of the current paper consists of the distinction between a cognitive stage (latent entrepreneurship) and a behavioral stage (actual involvement in entrepreneurship) and the link between both stages. This distinction has largely been absent in current female entrepreneurship research. It provides us with new insight into whether women’s lower level of activity in entrepreneurship is driven by a lower willingness and/or ability levels. Additionally, although distinguishing between mediation and moderation effects in the area of gender and entrepreneurship is not new (Collins-Dodd et al., 2004 and Verheul and Thurik, 2001), testing for such effects is novel within the present context: that of the latent and active stages of the entrepreneurial process. Our model for explaining entrepreneurial behavior is inspired by Ajzen’s (1991) Theory of Planned Behavior (TPB), which links behavioral intentions to actual behavior. In the next section, we introduce our conceptual framework. We then discuss gender differences with respect to the variables in our framework. The aim is to establish whether the impact of a variable on self-employment is different for women and men (moderation effect) and/or whether a variable has a different value for women and men (mediation effect). Subsequently, we introduce our model and discuss how we test for these gender effects. Finally, we present and discuss the results of the analysis and give suggestions for further research and policy. Recent data from 28 European countries, benchmarked with US data, guarantees not just the wide applicability of the results but also to detect specific country differences. Throughout the present paper, we will use the terms entrepreneurship and self-employment interchangeably.4

Motivated by relatively low self-employment preferences combined with a low self-employment prevalence rate for women, this study investigates the underlying mechanisms of this pervasive gender gap. Linking self-employment status to the preference for self-employment and using a representative data set of more than 8000 individuals across 29 countries, we examine different ways in which the gender dimension of the entrepreneurial personality influences the preference for and actual self-employment. These different ways relate to the inclusion of a cognitive stage (latent entrepreneurship), a behavioral stage (actual entrepreneurship) and the (perceived) ability to start a business in our model. Together, they offer a new way to contribute to the understanding of the origin of gender differences in entrepreneurial activity. We find a strong effect of gender on self-employment status through preferences: while the influence of preferences on actual involvement appears to be independent of gender (we find no interaction effect), women generally have a lower preference for self-employment. Next to this mediated effect through preferences, there is a significant ‘direct’ gender effect on actual self-employment (after controlling for preferences), indicating that, ceteris paribus, women have a lower chance of becoming self-employed than men. Lastly, looking at other factors included in the analysis, we find some support for moderation effects (interaction effects with gender). Taken together, our findings suggest that the relatively low self-employment rate of women is explained by both a relative lack of willingness and the existence of gender-specific obstacles, while these obstacles are felt more in the preference than in the action stage. The persistent and independent effect of gender (as measured by the coefficient of the gender dummy) on the preference for self-employment and particularly on actual involvement in self-employment suggests that there are other factors (than those included in the analysis) that are related to both self-employment and gender. For example, industry and entrepreneurial experience may have an important influence on the preference for and involvement in self-employment (Kolvereid, 1996a), but they are not included in this study as control variables. In addition, household and family responsibilities may play a role, where women simply feel that they lack the time to start a business. Moreover, women may feel that they lack the appropriate skills and knowledge for self-employment because they experience lower entrepreneurial self-perception (Ogbor, 2000 and Verheul et al., 2005). Thus, although preferences appear to be a key driver of the low self-employment rate among women, gender-specific obstacles may still exist. If their lower preferences keep women from starting up a business, it is important to understand where these preferences come from. We find several hints in our study. First, there is the mediating role of perceived behavioral control in the relationship between gender and entrepreneurial preferences. We see that women are less likely than men to believe that the economic climate is favorable for starting up a business and that they are more likely to believe that there are administrative complexities. Both perceptions lower women’s preferences for self-employment and, subsequently, their participation rate. In addition, the relatively low risk tolerance of women makes them less willing to become self-employed. The latter finding is consistent with Minniti et al. (2005), who find that fear of failure is important in explaining the lack of interest of women in self-employment. In terms of moderating effects, we find that assuming that an individual has self-employed parents, this has a smaller impact on women’s preferences than on those of men. It may be that men are more likely than women to be persuaded by parental role models when choosing a career. Alternatively, this may point at persisting traditional roles within families, where men are expected to take over the family business. The lower perceived behavioral control (or ability) of women, which underlies their lower preferences, is not driven by the belief that they are unable to control their own lives (an external locus of control) 37 but rather by a certain degree of pessimism regarding the outside environment and the extent to which it offers opportunities to start a business. In particular, the effect of perceived administrative complexity seems persistent for women, not only reducing their preferences but also (directly) discouraging them from taking action. This may be explained in terms of either real barriers or the perception of such barriers. Women may experience more administrative problems than men, for example, because they have less entrepreneurial experience or because they are active in industries characterized by high levels of red tape. It may also refer to a greater awareness of administrative procedures on the part of women. In this respect, it has been suggested that women are more realistic or less optimistic than men ( Niederle & Vesterlund, 2007). Nevertheless, when administrative complexity is perceived, this perception appears to have the same power to hinder women and men (i.e., there is no moderation effect). Thus, although women more often feel that there are administrative complexities than men, the hindering impact of this barrier is the same across gender. Despite the richness of the Eurobarometer data set, some drawbacks must be mentioned. First, we are unable to test for reverse causality. Particularly for perceptions, this may play a role because perceptions can be formed on the basis of experience with self-employment. Also, because both wage-employed and self-employed individuals expressed their preferences, this variable captures both the desire to be self-employed, for people who already run a business, and the desire to become self-employed, for people who do not run a business and desire to do so. Even though it is reasonable to assume that preferences influence actual self-employment status, as modeled in Eq. (2), we should be cautious in interpreting the relationship between preferences and actual self-employment. Indeed, an additional analysis controlling for reverse causality shows that latent entrepreneurship is partially endogenous (see Section 5.5). However, we argue that our conclusions are still valid once this reverse causality is controlled for. 38 Second, our data set covers a wide range of developed countries, and the results cannot be translated immediately to developing countries, particularly because it can be expected that preferences regarding self-employment (vis-à-vis wage employment) will differ between developed and developing countries ( Thurik, 2011). Third, as already stated above, industry and occupational experience, household and family responsibilities and detailed educational histories may contribute to the explanation of both preferences and actual self-employment. Other variables may relate to the different sector distribution of female entrepreneurial activities, geographical location ( Santarelli & Vivarelli, 2007) or the presence of actual (instead of preferred) financial constraints ( Alesina, Lotti, & Mistrulli, 2008). Given the untapped female entrepreneurial potential, it is important for policymakers to understand from where the gender differences in the perception of the entrepreneurial environment originate. This is a well-known gap in our understanding of the entrepreneurial personality. For example, is the greater perceived administrative complexity by women due to a real barrier – which would imply that a solution should be found to the red tape problem that women experience – or an awareness barrier – suggesting that women should be better informed of existing procedures and how to cope with them. In this respect, van Stel and Stunnenberg (2006) argue that governments should not only reduce the administrative burden but also communicate more clearly administrative procedures to potential entrepreneurs. Furthermore, it is possible that women are simply more pessimistic about environmental conditions and their own abilities. In this respect, Mitchell et al. (2002) pose that cognitive differences between women and men may be responsible for some of the unexplained gender differences. In general, government policy that is aimed at encouraging women to become entrepreneurs should not only focus on removing barriers but also address women’s preferences for and attitudes towards self-employment more directly. This may be done by providing information regarding the different type of risks involved in starting a business and how to cope with these risks or even offset some of them by acquiring relevant knowledge and skills. Also, paying attention to female role models may positively influence women’s self-employment preferences.