My mother-in-law does not like it: resources, social norms, and entrepreneurial intentions of women in an emerging economy

This paper provides new evidence at the intersectionality of gender, family status, and culture by focusing on a previously little researched group of middle-class women in an emerging economy. While the existing literature examines both structural and normative constraints for women entrepreneurship, little is known about the gains from relaxing structural constraints for women when compared to men. In addition to examining this new question, the paper sheds light on the binding nature of normative constraints for women entrepreneurship that persist in a patriarchal developing economy even when structural constraints are significantly eased. Using a mixed-methods approach, the empirical results suggest that higher resource availability differentially impacts the entrepreneurial intentions of women when compared to men indicating the strong presence of normative barriers that inhibit their entrepreneurship. These normative barriers emerge through the roles people play within women life spheres inhibiting their entrepreneurial intentions. Family members are holding back women entrepreneurship in patriarchal societies. Access to resources is not motivating women to become entrepreneurs as socially binding norms, and lack of support from family members such as mothers-in-law inhibits their entrepreneurial intentions. When policymakers in developing countries design programs to increase access to resources to future women entrepreneurs, they should address the normative barriers that discourage women entrepreneurship. In the absence of such support, gender inequities are likely to be exacerbated in developing countries. Women need to build resilience in the face of adversity and disapproval to pursue their entrepreneurship goals.


3
Vol:. (1234567890) role in enabling women entrepreneurship (Carter et al., 2001;De Bruin et al., 2007). The functionalist tradition assumes that women face disproportionate structural challenges in their resource mobilization efforts that severely limits their entrepreneurship prospects (Brush, 1992(Brush, , 2006Carter & Rosa, 1998;Lindvert et al., 2017;Welter et al., 2006). This has led to calls for additional enterprise support for women by boosting the availability of social and financial capital for supporting their entrepreneurial intention and behaviour (Calas et al., 2009;Leitch et al., 2018;Lindvert et al., 2017;Marlow & Patton, 2005).
Nevertheless, 'the-more-the-merrier' approach towards resources and enterprise support implementation is at odds with the increasingly contextualized nature of entrepreneurship (Grant & Perren, 2002). With male-dominated norms continuing to pose significant barriers for women at workplaces (Ford et al., 2020) and social norms having a negative impact on women's participation in the labour market (Naldini et al., 2016), the increasingly popular normative view considers entrepreneurship to be closely intertwined with issues related to gender, context, and institutions (Ahl & Marlow, 2012;Berg, 1997;Brush et al., 2009;Henry et al., 2016;Marlow & Martinez Dy, 2018;Meliou & Edwards, 2018).
In this context, little is known about the relative importance of structural versus normative constraints for women entrepreneurship in emerging economies. This paper examines if the intention to start a business develops equally for women and men when structural constraints related to financial accessibility and networks are absent. If it is not so, normative constraints inhibit women entrepreneurship, and addressing structural barriers alone may not be sufficient in encouraging women entrepreneurship.
For examining this, the study adopts a mixedmethods approach using data from Bangladesh, a country that is traditionally patriarchal. Using quantitative techniques on a database of 1780 respondents, we first estimate the impact of relaxing structural constraints for the entrepreneurial intentions of women. This quantitative analysis confirms that gains from easing structural constraints are significantly less for women when compared to men suggesting that normative barriers inhibit their entrepreneurship. Following this, in the second part of the analysis, we use qualitative analysis of interviews and focus-group responses amongst a selected group of middle-class women who report no structural constraints to examine the role of normative constraints for the entrepreneurial intentions of women. The novel insights emerging from the analysis suggest that normative pressures reinforced by close family members such as one's mother-in-law significantly inhibit the entrepreneurial intentions of women even when structural constraints are negligible.
The paper makes fundamental contributions to the current debates on women entrepreneurship by examining the role of context for the entrepreneurial intentions of women when compared to men. The empirical results presented here demonstrate that women experience significantly lower gains than men when structural constraints are alleviated in some contexts. Furthermore, by examining the relative importance of normative constraints, the paper offers novel perspectives on women entrepreneurship in a patriarchal emerging economy context.
The remainder of the paper is structured as follows: Section 2 reviews the literature on the role of structural and normative challenges in an emerging country context in shaping the entrepreneurial intentions of women. Section 3 presents the methodology and discusses the database. Section 4.1 presents the empirical results. Section 4.2 discusses the limitations of the quantitative results and provides an in-depth qualitative analysis of the barriers facing women in an emerging country. Section 5 presents the discussion and conclusion.

Theoretical background
2.1 Entrepreneurial intentions at the intersectionality of gender, family status, and culture: the structural approach An emerging body of scholarship has called for the widening of gender studies to go beyond the mere use of gender as a sorting device to take up an intersectional approach that enables diversity to exist between different female groups (Marlow & Martinez Dy, 2018). Increasingly, scholars view gender as a complex, multifaceted social construct that is created and developed through repeated interaction and enactment between individuals (West & Zimmerman, 1987). Thus, the associated meaning of gender varies depending on the cultural context, family background, and professional qualifications of a woman. Prior studies find evidence that structural barriers inhibit women's entrepreneurial intentions (Braches & Elliott, 2017;Henry et al., 2016;Marlow & Patton, 2005). Grounded in the resource-based view, the functionalists hold a uniform assumption on the impact of resource availability for entrepreneurial intentions and prospects. According to this view, the heterogeneity of resource possession plays a major role in the development of entrepreneurial intention and the eventual take-up of entrepreneurship (Alvarez & Busenitz, 2001;Barnett et al., 1994;Barney, 1991;Dierickx & Cool, 1989;Gulati, 1999;Hoang & Antoncic, 2003;Wright et al., 2012). The lower level of the entrepreneurial intentions of women is usually attributed to the structural inequalities between genders-that women have historically struggled to obtain resources and thus have needed to find support from different channels (Marlow & Patton, 2005).
Such support includes networking events with entrepreneurs and venture capitalists (Ekinsmyth, 2011) and training in financial literacy (Field et al., 2010) offered by national, international, and voluntary organizations (Carter & Jones-Evans, 2009), and structured mentoring schemes wherein males are partnered with females (Petridou, 2009). Enterprise support schemes are expected to provide women valuable insights into male working practices as well as access to resources (Godwin et al., 2006). These schemes assume that women can be placed on a level footing with men by equalizing resource availability between genders. The provision of financial and social capital is of increasing interest in scholarship in developing countries (Nichter & Goldmark, 2009) as entrepreneurial individuals rely on these resources in the absence of well-developed formal institutions (Harrison et al., 2018;Mair & Marti, 2009). Here, we focus specifically on financial accessibility and networks to examine if they enable women in a developing country context to develop entrepreneurial intentions.

Access to external finance
It is widely accepted that access to finance influences all stages of the business start-up process (Atherton, 2009;Cassar, 2004). While such access depends on the experience of an entrepreneur along with the legitimacy and the nature of offerings of a business firm (Le et al., 2006;Nofsinger & Wang, 2011), the perception of its availability also influences entrepreneurial intentions (Kwong et al., 2012;Robertson et al., 2003). In particular, the absence of external finance is found to discourage entrepreneurial endeavours (Cassar, 2004;Xiang et al., 2015).
Women face more obstacles in getting external finance in general as a result of gender stereotyping (Jayawarna et al., 2012;Marlow & Patton, 2005). For instance, Fay and Williams (1991) suggest that approvals for start-up loans for women are significantly lower than for their male counterparts. Women face significantly more obstacles in getting external finance, and this influences their entrepreneurial activity negatively (Roper & Scott, 2009). In developed countries, alternative external financial sources support women to start businesses (Hermes & Lensink, 2007;Khandker, 2005;Khavul, 2010) creating entrepreneurial ecosystems where women have greater opportunities to obtain start-up finance (Khavul, 2010). As a consequence, the functionalist view suggests that these women who do not report any problem in accessing finance are more likely to develop entrepreneurial intentions. We re-visit this scholarship by examining if relaxing financial constraints differentially impacts the entrepreneurial intentions of women when compared to men in a developing country context. In particular, relaxing financial constraints may not necessarily improve the entrepreneurial intentions of women in patriarchal societies, as strong normative barriers may inhibit their entrepreneurship. Thus, we hypothesise the following: Hypothesis 1: Relaxing financial constraints differentially impacts the entrepreneurial intentions of women when compared to men. In particular, women are less likely to have increased entrepreneurial intentions in comparison to men when financial constraints are relaxed.

Networks
Networks influence the entrepreneurial process (Jack et al., 2010;Yu et al., 2014) by providing access to essential resources to start a business (Adler & Kwon, 2002;Lockett et al., 2013;Xiao & Fan, 2014). They can positively influence self-efficacy and entrepreneurial intentions (Bacq et al., 2017) that may in turn influence entrepreneurial choice and success (Prieto et al., 2010;Semrau & Werner, 2014). The literature on networks is similar to the literature on access to finance in that they both acknowledge the presence of structural barriers. However, there is no conclusive evidence of networks being sufficient for enabling women to develop entrepreneurial intentions. In this context, 'inferior' quality networks-those that are homogeneous, familyand friend-orientated, or less growth-driven-have an impact on their entrepreneurial outcomes (McGregor & Tweed, 2002;Renzulli et al., 2000). However, as in the case of financial constraints, it is not clear if well-networked women develop similar levels of entrepreneurial intentions as men, particularly in an emerging economy setting. We extend this growing body of scholarship by examining if increased access to networks differentially impacts the entrepreneurial intentions of women when compared to men. Improved access to networks may not necessarily improve the entrepreneurial intentions of women in patriarchal societies, as the benefits of such access may not be realised if normative barriers within the family have inhibited women entrepreneurship in the past. Thus, we hypothesise the following: Hypothesis 2: Improving access to networks differentially impacts the entrepreneurial intentions of women when compared to men. In particular, women are less likely to have increased entrepreneurial intentions in comparison to men when access to networks is improved.

Household income
The connection between household income and entrepreneurship has been long established (Caputo & Dolinsky, 1998;Millán et al., 2012;Nikolaev & Wood, 2018;Smith, 2005). Extent studies suggest that individual household income can make a difference in their decision to be involved in entrepreneurship based on their ethnicity (Brynin et al., 2019;Smith, 2005). In a conservative developing country context, household income can be a significant determinant in shaping the entrepreneurial intentions of women.
In a patriarchal developing country context, men dominate the society and control household finance (Akhter & Cheng, 2020). In such cases, women have less financial freedom and have limited power to spend money. However, if household income is not sufficient to maintain the family, women may be allowed to start a business for additional household income. Unlike in developed countries, where household income has a significant positive influence on entrepreneurial decision and happiness (Bruce, 1999;Zhao et al., 2020), in a conservative developing country, low household income might encourage women to choose entrepreneurship to support the family. Thus, in the context of low family earnings, women might experience fewer normative barriers because of necessity-related reasons, and are likely to have increased entrepreneurial intentions. Otherwise, the normative barriers prevalent in a patriarchal society are likely to inhibit entrepreneurial intentions of women. Hence, we hypothesise the following: Hypothesis 3: Increase in household income differentially impacts the entrepreneurial intentions of women when compared to men. In particular, women are less likely to have increased entrepreneurial intentions in comparison to men when household income increases.
The conceptual framework outlining the hypotheses is provided in Fig. 1. The figure shows that in an emerging country context, access to finance, network, and household income significantly influences entrepreneurial intentions. However, these influences are different for women when compared to men.

Beyond structural barriers: the normative view
The normative perspective explains the fundamental non-structural reasons behind the inability of women to develop entrepreneurial intentions (Krueger et al., 2000). The underlying assumption here is that others' expectations and desires shape self-understanding while encouraging compliance with these expectations (Shapero & Sokol, 1982). Feminist interaction theories of status expectation and symbolic interaction suggest that women develop different selfexpectations regarding their ability to start a business because they are exposed to different institutional and contextual influences than men (Bird, 1989;Weick, 1995). These theories suggest that women's understanding of entrepreneurship is socially constructed (Blumer, 1986). Their interaction with others results in socialization that leads to common expectations (McPhail & Rexroat, 1979). Such socialization takes place through both direct and indirect interaction. Direct interaction arises by talking to each other (West & Zimmerman, 1987). Indirect interaction occurs through implicit mechanisms, such as tacit signalling and other mediums (Blumer, 1980;Engeström & Middleton, 1998). As interaction involves both receptivity and expression, unresolved discrepancies between the communicators result in conflict (Gutek et al., 1991). Such interactions become particularly complex when they are embedded within a wider power discourse (Kibria, 1990). Within the context of working women, a fine balance is often established between work and family through an intricate connection of norms, expectations, and power relationships between multiple and interdependent actors (Duxbury et al., 1994;Gutek et al., 1991). Female participation in entrepreneurship can challenge the established power structures within a family, unsettling the previously established distribution of responsibilities and activities. Any small move can generate multiple changes in the power structures of the household.
Studies have previously explored the power struggle faced by women, which impedes their ability to start up (Bruni et al., 2004;Marlow & McAdam, 2013). The theory of power ritual suggests that powerholders can often repress subordinates through various mechanisms (Goss et al., 2011). The struggle that women face for participating in entrepreneurship is often described as a continuous battle (Heilman & Chen, 2003;Koutsou et al., 2003;Ragins & Winkel, 2011). Multiple normative battlefields, including social and emotional (Ragins & Winkel, 2011), discursive (Leffler, 2009), and those revolving around expectations of (in) ability or (in) competence (Heilman & Chen, 2003), characterize women's entrepreneurial endeavours with interactions resulting in multiple battles. These battles are most likely to be lost when they are most prone to disrupt the core of the power structure or when the rejection comes directly from significant power wielders within their social context. Nevertheless, women can propose compensatory or counter rituals, but these can disrupt other parts of the status quo, which may be preferred by women (Goss et al., 2011). What is unclear in the literature is the role of these normative barriers when structural barriers are absent. We make a novel contribution to the entrepreneurship literature by examining the role of such normative pressures in shaping entrepreneurial intentions of relatively more educated, well-to-do women who are not limited by significant structural constraints in an emerging economy context.

Methods
We use a mixed-method approach for examining the functionalist hypotheses and the normative barriers to women's entrepreneurial intentions in the Bangladeshi context.

Sampling and data collection
The quantitative data were collected from 2000 respondents through an extensive random field survey in all seven administrative divisions of Bangladesh in 2011. Respondents were selected randomly from

Dependent and independent measures
A summary of the variables used in this study is given in Table 1. The dependent variable is Entrepreneurial Intention, where we asked the respondents how likely it was that they would like to start a business within the next two years, with a 7-point Likert scale (7=most likely, 1=least likely). We have three independent variables: access to external finance, networks, and household income (as a proxy to family status). As we are interested in examining whether the effect of our independent variables will be uniform as the functionalists suggested, we examined the moderating effect of being female (1=female) on these variables in entrepreneurial intention. For Household Income, respondents were asked to record the income of all members of the family in one of the seven categories. To measure their network, we asked respondents to state, on a 7-point Likert scale, the extent to which their friends, family members, and other people are known to them could help them start a business (7=most likely, 1=least likely). We found statistically significant differences between female and male responses to this (3.81 as opposed to 4.45). To measure the perceived access to external finance, we asked respondents to state, on a 7-point Likert scale, the extent to which they would be able to collect funds to start a business from sources other than from their family (or other personal sources), if necessary. There are significant differences between female and male responses (4.08 compared to 4.83). Finally, for control variables, age, education, household size, parents' entrepreneurial background, perceived skill, attitude towards selfemployment, opportunity recognition, fear of failure, and social status are included in the estimations. Here, age is the exact age of the respondents.

Qualitative methods for normative analysis
In the qualitative part of the study, we conducted indepth interviews with 18 women. A key part of the research strategy is that they did not report notable problems in accessing finance and networks but nevertheless have considerable reservations in terms of whether they can start a business in the immediate future (within two years). As far as we are aware, such a female sub-group has rarely been looked at, as the focus of most studies is on the poor, who face significant daily-life struggles and are driven by the necessity to find ways to bring income to the family. For analyzing these interviews, we use a qualitative social network analysis approach (Heath et al., 2009). At first, the interviewees are asked about their views on entrepreneurship and their perceptions of the functional and normative obstacles to entrepreneurship. Following this, they reflect on how their middleclass upbringing has positively or negatively impacted their entrepreneurial intentions. Then they define the egocentric network, specifically highlighting the key actors who may potentially contribute to their entrepreneurial endeavours and attributing the sources of the perceived functional and normative obstacles in relation to these actors. Finally, the interviewees are asked about the role of each of these actors in affecting their entrepreneurial intentions and prospects. Table 2 presents the regression results 1 to estimate the impact of structural constraints on entrepreneurial intentions. In Model 1, the baseline specification with the control variables suggests that skills, opportunity recognition, parents' entrepreneurial background, age, and household size have a significantly positive 1 We used ordered logit model as the entrepreneurial intention is measured here on an ordinal Likert scale variable. To test the existence of multicollinearity, we used the variable inflation factor (VIF). Any value of VIF above 10 indicates the possibility of the existence of multicollinearity. Since the value of VIF is less than 4, there is no possibility of multicollinearity. A tolerance value of more than .1 also indicates that there is no multicollinearity. influence on entrepreneurial intention while the influence of fear of failure is negative and significant. Attitude towards self-employment, education, and social status are insignificant. Education and social status have negative signs, suggesting that individuals highly endowed in this might either have better opportunities in employment or simply no need to start up a business themselves. Model 2 introduces gender along with the independent variables. Being a female is found to have a significantly negative (p<0.01) influence on entrepreneurial intention. It also shows that access to finance, networks, and household wealth has a significantly positive influence on entrepreneurial intention. These are consistent with our expectations. The control variables mentioned above remain largely unchanged.

Empirical results for the functionalist view
Models 3-5 introduce the interaction terms of gender with access to finance, networks, and household income separately. Because of the introduction of the interaction terms, the main effects on all of the variables diminished as expected. The interaction term between gender and finance, networks, and household income variables are all significantly negative. As the entrepreneurial intention is positively correlated with the perception of financial availability, the negative interaction term suggests that there is an additional negative impact of being a female. This implies that the effect of finance on encouraging entrepreneurial intention is smaller for women than for their male counterparts. Therefore, women with access to finance do not necessarily develop entrepreneurial intention as much as their male counterparts. Similarly, the negative interaction effect between gender and networks suggests that although networks positively impact the development of entrepreneurial intention, women's entrepreneurial intentions do not benefit as much as men's from access to networks. Finally, the results suggest that females from households with more income do not develop entrepreneurial intentions as much as their male counterparts. These striking results support the core conjecture that relaxing structural constraints have a differential and lower impact on the entrepreneurial intentions of women when compared to men. Figures 2, 3 and 4 explain the marginal effect of the interactions. In the figures, the red lines are the predicted probabilities for women, and the blue lines are predicted probabilities for men. The blue lines in Figs. 2a and 3a demonstrate that, with perceived increased access to external finance and networks, the disagreement to having entrepreneurial intentions decreases for men while it is not so for women. In Figs. 2b and 3b, the red lines are flat in all three interactions suggesting that increased access to finance and networks does not change the probability of the entrepreneurial intentions of women.
In the case of household income, women are less likely to pursue entrepreneurship than men if household income increases (Fig. 4). The gap between men and women widens with growing differences in household income.
These results lead us to infer that normative barriers to female entrepreneurship may explain the fact that the gender gap persists even when structural barriers to the entrepreneurial intentions of women are addressed. These compelling results highlight that the problem of gender differences in entrepreneurial activities arises not only because of structural inequality in resources but also because women tend to gain less when these structural constraints are eased, suggesting that normative barriers play a significant role in shaping entrepreneurial intentions of women. These are examined in the following section.  3.782*** 3.271*** 3.534*** 3.660*** 3.806*** 4.052*** Standard errors in parentheses ***p<0.01, **p<0.05, *p<0.1

Qualitative results for the normative view
The sample used here represents a sub-group of women that is distinct from previous studies. The women in our sample are well-educated and consider themselves as belonging to 'middle' or 'upper middle' classes within the society (Table 3). We expect that a different power structure may exist in their case in comparison to those found within the extant literature. For example, their education and relative financial stability should produce an empowering effect, as they provide more choices and options to them. Furthermore, while this sub-group of women interviewees accept that the 'purdah' system may hinder entrepreneurship participation amongst Bangladeshi women, some of them felt that they were significantly bound by it. The women in our sample do not see interaction with outsiders as a problem. This suggests that their education and upbringing put them through a socialization process that is different from that found in traditional studies that focus on poor women. This results in them having a diverse and significant social network that can potentially support them. Their networks have entrepreneurs, government officials, and financiers who can provide resource support and act as mentors. While recognizing that there are existing structural problems for them in accessing finance, they felt that they had a good understanding of the procedures and critical thinking skills that are needed to overcome the structural problems as and when they arise. Furthermore, they are aware of the different sources of support that they can access, including formal finance and other sources of funding (e.g. NGOs and other loans mentioned in transcriptions). In particular, some have pointed to sources of finance that are only available to women.

Fig. 4. Household income and gender interaction effects
Although not intending to start a business within the next 2 years, most women reported having a positive image of entrepreneurship, and some recognised the empowering effect that entrepreneurship could have on women. Despite this and the fact that resources and networks are not a problem for them, these women still did not develop their entrepreneurial intentions.
We find that most women within this sub-group perceive significant normative, as opposed to structural, barriers towards entrepreneurship. The qualitative analysis suggests that there are multiple pressures, oppositions, and challenges and that women have to engage in multiple battlefields if they are to contemplate entrepreneurship as a career path. In the following section, we highlight a number of players within the power structure and their views about the participation of women.

The family system
Actors within the family system play a crucial role in the shaping of the meaning of entrepreneurship. Many had strong views about women's participation in entrepreneurship, although their influence on women was of varying degrees. For those who are single, their family is highly influential, while, for those who are married, the spouse's family-in particular, the mother-in-law-plays a more central part within the power structure. Our study discovers a number of mechanisms by which pressures were being exerted on the women respondents.

Disapproval, deflation, and sabotage
Family members that 'disapprove' of the woman's entrepreneurial activity are particularly concerned with how much activity reflects on their family's standing and even suggest that such activity 'brought shame to the family' (C1a , Table 4). Some husbands particularly resisted the idea as this 'reflects badly' on them personally. Families often adopt the strategy of 'deflation' (C1b, Table 4). One respondent even imagines that her family would 'sabotage' her attempt, as they did with her previous employment (C1c , Table 4).
Conditional support Women found some family members to be 'encouraging' (C1d, Table 4). Generally, such encouragement comes from immediate family members who have a strong influence on the vision and values held by the women. Some respondents also reported that their husbands were 'supportive'. Nevertheless, they found such encouragement or support to be 'conditional', depending on the type of business-typically small and requiring minimal interaction with strangers. Furthermore, they had to prioritize housework even in cases where entrepreneurial activity is encouraged.

Withheld permission
The husband, or the in-laws, are concerned with the women's entrepreneurship choice since it affects their daily activities. As a family unit, they hold power to 'withhold permission', creating institutional barriers to participation (C1e, Table 4). The resistance comes mainly from the other women within the household, led by the motherin-law, as they have to take on additional domestic duties.
Responses from women towards family pressure Women find it particularly challenging to navigate between the contradictory ways they portray their family: on the one hand, they often bestow them with 'enlightened middle-class respectability'; on the other hand, they identify the 'backward view' held by the family towards female entrepreneurship. This creates considerable 'disappointment', 'frustration', and 'anger'. Consistent with the symbolic interaction literature, most women have chosen to 'maintain harmony' (R1a, Table 4), and left such a view unchallenged to avoid further ramifications from the collapse of the power structure. Others chose to 'compromise' (R1b , Table 4) by lowering their expectation of entrepreneurial participation. Some favoured boutiques and beauty parlours in order to conform to the norm while avoiding sectors such as real estate, construction, and transportation. This is consistent with the previous literature on symbolic interaction, which observes that, as women realize that the meaning of entrepreneurship has changed, they adjust their expectations accordingly. However, some are reluctant to 'conform' (R1c , Table 4), as the prospects of legitimate businesses tend to be unattractive, while the non-committal stance could hinder the business outcomes. Another strategy raised was to 'negotiate' (R1d ,  Table 4). However, alongside a desire to develop an entrepreneurial career that would give them income and autonomy, the women also desire other identities, in particular, as a respectful wife, mother, daughter (in-law), and citizen of their class. As the women are conscious of the delicate balance between the different bargaining chips they had to hand, some said they would 'calculatedly opt-out' (E1e , Table 4), either by de-prioritizing entrepreneurship or by delaying participation until a much later stage.

Non-business-related influences
Upholding social norms Women have reported that community elders and other senior citizens felt that it was their duty to 'uphold the societal norms' (C2a , Table 5). They express disapproval of women's entrepreneurial participation, citing 'religion' and the 'irresponsibility of neglecting their family duties.
In response, some women demonstrate a 'rebellious' streak towards these established views, considering them to be old-fashioned and narrow-minded (R2a , Table 5)-some disputed, in private, their religious grounds. Consistent with Essers and Benschop (2009), they felt that their educational privilege enabled them to criticize the 'wrong interpretation', 'inconsistencies', and 'inaccuracies' of Muslim teachings by these elders. Their elder's disapproval had a little direct impact, as they lived in the 'modern segment of the society'. Although the older generation is less educated and more isolated from modern society, they still represent the norms of society and, as such, have an influence on the mindset and attitudes of family members. The women feel they are 'indirectly confined' by this influence (C2c , Table 5).
Gossip Many women in the sample point out that it is common for bystanders to 'gossip' (C2b , Table 5) and create a bad image of female entrepreneurs and their families. While the gossipers made a little direct impact, the women found them 'hard to ignore', as they are often closely connected to the women's family members through socialization and business dealings (R2b , Table 5).

Business-related influences
Face value support, subtle discouragement Many women in the sample report that they knew entrepreneurs in their network who can help them start a business. Most of these entrepreneurs in their networks are women who initially encouraged others to start businesses. Some of these entrepreneurs promised both financial as well as non-financial support to women who intended to start up. However, when some of the women in the sample decided to actively pursue this further, the responses turned into 'subtle discouragement' (C3a, Table 5). While this shift intrigued some of the women in the sample, others felt that the tepid responses might be due to women having limited choices compelling the entrepreneurs to avoid actively encouraging competitors to enter an already intensely competitive market environment. While the women accepted the responses as understandable, some felt disappointed by such subtle discouragement that instills 'fear' (R3a, Table 5) and puts them off from starting a business. Figure 5 provides a summary of the different actors involved, the contrasting views, and how these battles were fought by the women. It presents the normative pressures, influences, and the corresponding responses of women. The green arrows show positive influences, the red arrow shows negative pressures, and the blue arrows show women's response to these influences and pressures. The width of these arrows stands for the strength of the influence or response. Non-business external actors such as local leaders and enforcers of social norms that include community elders and senior citizens uphold societal norms and, in the process, impose normative barriers on middle-class women entrepreneurship. This is given by the red arrow from External Actors (non-business-related) to Woman in Fig. 5. The women from middle-income groups can offer a strong rebellious response or ignore such imposition of norms, as such non-business-related external actors do not come in direct contact with the women. This is given by the thick blue arrow that is directed from Woman to the External Actors (non-business-related) in Fig. 5. However, when the members of their family become mediators for the imposition of such societal norms (given by the thick red arrow on the left of Fig. 5), and the family exerts strong negative pressures on the women in a patriarchal society, they cannot ignore such norm enforcement and tend to compromise their entrepreneurial intentions. Similarly, the red arrows on the right side of Fig. 5 suggest that when external business actors, particularly women, talk about their past negative experiences and challenges in business, they create subtle discouragement for women's entrepreneurial intentions. Any encouragement from them becomes a nominal exercise that has a weak positive influence on women's entrepreneurial intentions.

Discussions and conclusion
The normative masculine discourse of entrepreneurship is highly pervasive (Ahl, 2006;Marlow & Martinez Dy, 2018), while external context is believed to have a significant influence on the development of entrepreneurial intention (Welter, 2010). Furthermore, men and women have different interpretations of reality, time, action, interaction, power, and ethics, all of which may affect the way they identify opportunities or barriers (Bird & Brush, 2002;Braches & Elliott, 2017;DeTienne & Chandler, 2007). Our study makes novel contributions to the literature by focusing on a group at the intersectionality of gender, family status, and culture that rarely received attention in extant scholarship. We highlight the challenges and pressures faced by middleclass women living in the patriarchal, developing nation of Bangladesh when trying to launch a business venture. Through a mixed-method approach, the paper examined how structural and normative challenges affect women's entrepreneurial intentions. While the structural and normative challenges that women face in developing entrepreneurial intention has been studied (Baughn et al., 2006;Marlow & Patton, 2005)-especially in the developing countries context (Jamali, 2009;Roomi & Parrott, 2008)-this paper is one of the first systematic studies exploring how relaxing structural challenges affects the entrepreneurial intention of two genders differently. The compelling results presented here suggest that when structural barriers are relaxed, women remain much less likely than men to develop entrepreneurial intentions. Thus, the quantitative part of the study highlights the fact that a lack of structural challenges to starting a business does not result in increased entrepreneurial intention for women, which suggests that entrepreneurship in this particular group is impaired by other, potentially normative, issues.
With this result established, the qualitative part of the results uncovers the normative mechanisms that affect the development of their entrepreneurial intentions. In particular, the paper focuses on a subgroup of women who are relatively unhindered by structural constraints to examine other issues that affect their intention to start a business. The novel insights emerging from here suggest that many of them indeed see entrepreneurship in a positive light and expect participation in entrepreneurship to be an empowering experience. However, the complexity of constructing a positive entrepreneurial identity for the women in this sub-group inhibits the development of their entrepreneurial intentions.
This process is different from the one experienced by women of poorer family backgrounds that are explored in the extant literature (Harrison et al., 2018;Mair & Marti, 2009). In their cases, the barriers are mostly structural. For those living in poverty, starting a business is not normally an aspiration but often a necessity, as family pressures push them towards entrepreneurship. Structural constraints compel this group to engage in subsistence, low growth entrepreneurial endeavours, which is the alternative outcome of no income at all. These types of entrepreneurial endeavours are deemed acceptable to the patriarchal society because they do not threaten the existing power structure or status quo but provide additional family support. However, women from middle-class groups are reluctant to engage in such necessity entrepreneurship. Thus, a core theoretical contribution of this paper is to highlight how the lack of entrepreneurial intent for women without structural barriers arises from a number of normative challenges. We contribute to the literature by highlighting some key mechanisms that manifest in women's interactions within Bangladeshi society (Tables 4 and 5). These normative barriers came from different spheres of a woman's life wherein actors often mutually reinforce the discourse that female participation in entrepreneurship should be minimal. While the finding is specific to the Bangladeshi context, many of the findings can be extended to other similar societies, where informal, social, and cultural institutions play a key role in shaping norms and expectations and where a strong male-dominated discourse exists.
The insights presented here are consistent with previous studies from the employment literature, which suggest that women deliberately set their sights on homemaking, and, time permitting, an employment path that offers flexibility for family duties, resulting in part-time, contracted work from home that has fewer career prospects (Anna et al., 2000). We also find that even if women cross the hurdles of entering entrepreneurship, there are limited to entering specific sectors with entrepreneurship opportunities clearly divided into 'permitted' and 'forbidden' categories. While the former offer flexibility for housework and require little interaction with outsiders, they also tend to have low profit and growth potential. The respondents are well aware of the options available to them. These results are broadly consistent with emerging research which suggests that women drift into selfemployment, but only towards the flexible, low-end types with low growth potential (Thompson et al., 2009). These cases illustrate that women are pushed into entrepreneurship out of necessity or desperation, but not necessarily because a positive entrepreneurial identity has been developed (Bryant, 2013). Such entrepreneurship is deemed relevant and acceptable for those considered to be belonging to lower rungs of the socio-economic class hierarchy but not necessarily appropriate for middle-class women. As a positive entrepreneurial identity is not developed amongst the middle-class group, their access to finance and networks does not necessarily increase their intention to start new ventures. Where women may have desired to start a business within these 'forbidden' sectors, the perceived unwelcoming attitudes amongst the formal and informal institutions cause fear and despair, pushing back their entrepreneurial ambition, leading them to opt-out of the unattractive prospect of entrepreneurship altogether. The result that women in our sub-group find the idea of entrepreneurship to be positive and empowering and that they are interested in considering entrepreneurship more seriously if they are unhindered by social and cultural institutions suggests that there are pertinent policy and managerial implications to these findings.
The compelling results presented here also have implications beyond developing country contexts. Increasingly, scholarship on women entrepreneurship is attempting to explain the significantly low numbers of women entrepreneurs in developed countries . Although it is widely recognised that women entrepreneurship has a positive impact on the economy (Bullough et al., 2019), similar normative barriers can potentially inhibit their entrepreneurship (Field et al., 2010). Such barriers include gender-based role expectations , Bullough et al., 2021. Furthermore, women from ethnic minority communities in developed countries may continue to face both structural as well as normative barriers although they are living there for a long time (Audretsch et al., 2017;Brynin et al., 2019;Carter et al., 2015)-an effect that can be attributed to the persistence and sticky nature of these barriers. Such women are more likely to be pushed into entrepreneurship with low returns (Baycan-Levent & Nijkamp, 2011). Thus, transcending structural barriers to women entrepreneurship alone may not be sufficient in promoting their entrepreneurial intentions if the normative barriers persist in different forms across varied contexts (Baughn et al., 2006).

Policy implications
Despite the increasing emphasis on the role of social and cultural institutions in the pursuit of new ventures, the majority of policies continue to focus on eliminating structural challenges for those intending to start up, through providing resources and network support. The findings here raise compelling questions regarding this inherently structural-driven approach. We find that although addressing structural barriers improves the entrepreneurial intentions for men, they may be not sufficient for supporting women, particularly in the presence of non-structural normative challenges that affect women's entrepreneurial intentions. We argue that while policymakers should continue to offer structural support for both genders, they should also address the normative barriers that hinder female respondents. While not all normative challenges can be overcome, and some of the changes require time to take effect, we propose that policy can begin to tackle some of the issues at individual, legislative, and societal levels.
At the individual level, the policy can help reduce the burden on women in the domestic arena, for instance, by finding cost-effective solutions to family care, targeting both pre-school age and after-school care for those of school age, as well as the elderly and the sick (Ahl, 2006;Blackburn & Ram, 2006;Rouse & Kitching, 2006). In the context of developing countries, where institutional voids are present, developing a platform to enable a shared economy to prosper, and hence allowing the sharing of these family duties, can be more cost-effective (Kirkwood & Tootell, 2008). At the legislative level, the government should consider policies that will reduce the reliance on normative-based social and cultural institutions. For instance, if gaining family permission to start a business is one such stumbling block as many of the women in the sample mentioned, advancing legislation to minimize the requirements for such permission can unleash their entrepreneurial endeavours. At the societal level, the government faces a mammoth task to change the overall attitude, particularly amongst the middle class, in terms of female entrepreneurial activity. The interviewees amongst the targeted sub-group suggest that there is a discrepancy between their personal perception of entrepreneurship and how it is treated by people within these women's spheres of life. Their inability to develop entrepreneurial intention is largely attributable to the latter. While they believed in their ability to overcome the negative views of outsiders at a personal level, they have significant concerns about how others may judge their families due to their participation in entrepreneurship, as this may upset their family members. They have put off entrepreneurship because they felt the pressure to maintain harmony at home.
Since these long-standing norms are well embedded in both formal and informal institutions, overturning them would be impossible unless one can, as Calas et al. (2009) suggests, carry out extensive socialization through familial and educational institutions. As in previous studies, these norms and expectations are being upheld and supported by various formal and informal establishments (DiMaggio, 1988;DiMaggio & Powell, 1991;North, 1990), whose views are often rejected by the particular subgroup that we studied, but nevertheless remain dominant in the society at large. For instance, with the role of religion in female entrepreneurship participation remaining debatable, as suggested both by the respondents and in the academic literature (Essers & Benschop, 2009), there is a strong need to open up the debate and provide channels for the expression of non-dominant discourses. Policymakers can initiate forums for both sides to express their views regarding the dominant discourse that is perceived to be inaccurate by the women (Gray & Finley-Hervey, 2005). There is also a need to promote role models, in the form of successful female entrepreneurs, by offering them official roles to disseminate information about the advantages of entrepreneurship amongst women in the public spheres (Bosma et al., 2012). Finally, policymakers can sponsor media events to promote female entrepreneurship (Carter et al., 2006). Only by slowly changing the mindset of the society towards entrepreneurship can these women choose to participate in entrepreneurship with fewer inhibitions that arise from fear and anxiety.

Managerial implications
Although the women in our targeted sub-group did not express an intention to start a business in the near future, they are not uninterested in entrepreneurship, and indeed many regarded entrepreneurship as an empowering notion. This has two implications. First of all, that normative social and cultural expectations do indeed create further practical challenges to starting a business, and second, those normative challenges create fear and anxiety that reduce entrepreneurship participation. To tackle the normative priority of home overwork, women need to think creatively in terms of how traditional household duties can be reorganized (Biernat & Wortman, 1991). There is also scope for bottom-up collaborations between budding entrepreneurs, for instance, through sharing childcare, school drop-off/pick-up, commune style cooking arrangements, and other possible dutysharing arrangements which can free up their time to engage in entrepreneurial endeavours (Wilson et al., 1990). Much of these are already happening in various parts of the world, and in the densely populated context of Bangladesh, such arrangements can be replicated. There is also a question of negotiation as well as developing strategic alliances with those who can offer them support (Yeoh & Willis, 2005).
Finally, while it is not considered ideal for women in the upper-middle and middle classes to participate in entrepreneurship, as gender clearly intertwines with class in the process of symbolic interaction, women should not have to feel disheartened by the situation. Unfortunately, some of the respondents who described entrepreneurship as their dream apparently did feel disheartened because of these normative pressures. There is a need to develop resilience in the face of adversity and disapproval and to pursue their entrepreneurship goals against all odds. In turn, the women themselves can be the game changers, as more of them become successful and influential and are not confined by social norms.