What drives MSME performance? The role of gender

This study aims to investigate the effect of gender toward the ease of access to capital, entrepreneurial behaviour, and the performance of MSMEs. In contrast to previous studies, this research is also testing the moderating effect of the patriarchy on the relationship between gender and the variables of interest. Data were collected using a self-administered questionnaire distributed to 201 MSME entrepreneurs throughout Bali Province of Indonesia and analysed using a partial least square structural equation modelling (PLS-SEM) technique. Results indicated that (1) female entrepreneurs of MSMEs have a lower level of perceived ease of access to capital; (2) patriarchy positively moderates the relationship between gender and the ease of access to capital; (3) female entrepreneurs of MSMEs have lower performance than its counterparts; (4) entrepreneurial behaviour has a positive effect on MSMEs performance; and (5) the ease of access to capital have a positive impact on entrepreneurial behaviour. This study suggests the potential moderating role of patriarchy on the relationship between gender and the ease of access to the capital of MSMEs. The analysis of the business capital environment situation in Indonesia conducted by Rachmina (2009) shows that the existence of patriarchal culture has limited women's accessibility to capital programs organized by the government. In terms of credit policy, several programs are found to be biassed towards female target groups, such as the Urban Poverty Reduction Program ( Program Penanggulangan Kemisikinan Perkotaan — P2KP), Independent Business Performance Program ( Karya Usaha Mandiri — KUM), and the Small Farmers and Fishermen Income Improvement Program ( Program Peningkatan Pendapatan. Petani-Nelayan Kecil — P4K). It is due to the general idea that under competitive conditions, the level of potential for success in women-run businesses is relatively lower than that of men. Likewise, the patriarchal environment has made women perceived as more appropriate to act as managers of household affairs, and only men have the right and obligation to work to meet their family needs. Consequently, the number of women who have businesses is found to be relatively lower, while in terms of credit accessibility women entrepreneurs who manage to obtain loans are also in a smaller proportion than men (Rachmina, 2009). This study proposes the following hypothesis:


Introduction
Amid the increasing number of MSME business units in Indonesia 1 , statistical data shows that 23% of MSMEs are owned by women with a business growth enhancement of 8.1% per year (The Asia Foundation, 2013). Meanwhile, the number of MSMEs owned by men was found to be decreasing. In addition, the proportion of female entrepreneurs of MSMEs (from now on, FMSMEs) who reported achieving business profitability in the "very satisfactory" category was also found to be higher (41%) compared to male entrepreneurs of MSMEs MSMEs (from now on, MMSMEs) (31%). In 2013, FMSMEs was noted to have contributed positively to the Indonesian economy in the amount equivalent to 9.1% of the country's Gross Domestic Product (GDP), which amounted to 443 trillion rupiahs for small businesses and 421 trillion rupiahs for medium-sized companies (International Finance Corporation -World Bank, 2016).
FMSMEs performance in Indonesia, unfortunately, is still overshadowed by the issue of gender equality which is related to, amongst others, the ease of access to capital in running a business. Surveys conducted by the International Finance Corporation and World Bank in 2016 found that the proportion of women who perceived the complexity of the procedure for granting bank loans as a negative factor for the survival of their MSME businesses was higher (40%) compared to men (28%). In addition, Kholifah (2011) also found that women have a lower ability than men in terms of access to capital assistance from government programs. In some previous studies Entrepreneurial behaviour displayed between individuals can be very different since it is closely related to the differences in demographic characteristics possessed, such as age, gender, and education level. In a gender perspective, differences in entrepreneurial behaviour can arise as a consequence of the different forms of "social credential" given by the environment to men and women (Acheampong, 2018). Socially, men are more "trusted" to be able to engage in labor market activities, while women are more "trusted" to be able to manage household tasks. This social arrangement further influences how the community environment negotiates and allocates social resources differently to men and women. Meanwhile, those intended social resources are the primary "raw materials" for the success of the process of creativity, learning, and innovation carried out by an entrepreneur (Sarasvathy, Kumar, York, & Bhagavatula, 2014). Verheul and Thurik (2001) reveal that men and women display some differences in their entrepreneurial characteristics. Men are found to be giving more appreciate on the aspects of perseverance and courage in taking risks, while women in issues of independence and change. Interestingly, women tend to perceive lower entrepreneurial capacity than men (Verheul & Thurik, 2001). Alma (2013) states that emotionally, women are more lenient and compromised in facing the mistakes of work partners or employees and are naturally limited in space and time by their inherent household obligations. On the other hand, Sherlywati et al. (2017) found that MMSMEs were more confident in facing competition challenges, had a more long-term orientation in business development and were more flexible and adaptive to changes that occurred in their environment. Based on the above description, this study proposes the following hypothesis: H1: Gender affects entrepreneurial behaviour of MSME entrepreneurs, in which MMSMEs display more superior entrepreneurial behaviour than FMSMEs Gender, Entrepreneurial Behaviour, and Patriarchy Gray (2002) states that entrepreneurial behaviour is a variable related to specific situational contexts. In other words, the location, industrial environment, and organizational size must be taken into consideration when identifying entrepreneurial behaviour s in carrying out a business cycle. Kirkley (2016) also emphasizes that entrepreneurial behaviour can be seen as a form of behaviour based on social and cultural aspects and is the result of interaction between individuals, society, and the surrounding culture. Entrepreneurial behaviour will arise when there is the proper carrying capacity of the surrounding community. This support can be derived from the interpretation of the social environment that entrepreneurial behaviour s have been congruent with the values and norms that are widely adopted by the relevant community groups (Kirkley, 2016). Verheul and Thurik (2001) revealed that there exist differences in entrepreneurial characteristics possessed by men and women. For example, regarding the risk acceptance attitude, women tend to perceive their capabilities to be lower than those of men. This negative attitudes can then be facilitated by social and cultural factors, such as the subordination of women's roles in life and the internalization of human values and images in a gender perspective. Meanwhile, several other researchers have proven that aspects of entrepreneurial behaviour are found to be higher in the community environment with a high level of masculinity culture (Hayton & Cacciotti, 2013;Stewart, Carland, Carland, Watson, & Sweo, 2003;Urban, 2006;Urban & Ratsimanetrimanana, 2015). Masculinity is a cultural dimension in a society where the individual's social roles based on gender are clearly defined-in which men are assertive, resilient and oriented towards material success, while women are more humble, sensitive, and concerned with the quality of life (Hofstede, 2001).
This study suggests the potential moderating role of patriarchy on the relationship between gender and entrepreneurial behaviour of MSMEs entrepreneurs. Identical to the dimensions of cultural masculinity initiated by Hofstede (2001), patriarchy is also a social aspect that is closely related to a gender perspective. This culture refers to the universal values that apply and is accepted in society for the existence of structural control that is more dominantly owned by men-relative to women-on political, legal, economic and religious aspects (Sakalli-Ugurlu & Beydogan, 2010). A patriarchal community environment positions women in domestic roles, while on the other hand, associates men with careers, authority, and leadership. This social "limitation" then has the consequence of the low motivation of women to bring up and display entrepreneurial characters they have. In the Indonesian context, the results of Kholifah (2011) study have revealed the negative impact of patriarchal culture on the resilience of women MSMEs in facing business risks, towards self-confidence and towards business development efforts. Based on the above description, this study proposes the following hypothesis: H2: Patriarchy positively moderates the effect of gender on entrepreneurial behaviour, in which the superiority of entrepreneurial behaviour displayed by MMSMEs are getting higher under the conditions of the high level of patriarchy

Gender and the Ease of Access to Capital
MSMEs face challenges in terms of ease of accessing the capital (financial resources) from external parties (European Central Bank, 2017). The term "ease of accessing capital" refers to the level of ease of obtaining information, services, and approval of the application for funds (capital) submitted. This phenomenon is quite apprehensive by considering that the development and enhancement of MSMEs potential growth need actual existence of a conducive business environment, primarily related to access to funding (International Finance Corporation -World Bank, 2016). Financial resources are crucial because basically all forms of profit-oriented organization, including MSMEs, require capital to be able to carry out their operational activities. Although MSMEs initially were able to operate by maximizing the potential of financial resources originating from internal (own capital), as the business progressed MSMEs could not avoid the need to obtain additional funding from external parties (Verheul & Thurik, 2001). Interestingly, the level of ease of access to capital can be perceived differently by male and female entrepreneurs of MSMEs. Some researchers (Ardhanari, 2007;Coleman, 2000;Kholifah, 2011;Muravyev, Talavera, & Schäfer, 2009;Nainggolan, 2016;Nguyen & Nguyen, 2014) found that FMSMEs tend to feel a lower level of ease in accessing financial resources from external parties. The business environment analysis in 34 countries worldwide by Muravyev et al. (2009) revealed that MSMEs managed by women have a smaller chance of getting loans and are more likely to be burdened with higher interest rates than MSMEs managed by men. Likewise, in the aggregate, the amount of credit approved for FMSMEs was also found to be smaller than for MMSMEs (Nguyen & Nguyen, 2014). Coleman (2000) observed the implementation of stricter evaluation procedures for loan applications submitted by FMSMEs by banks and other financial institutions. MSMEs managed by women are perceived as having relatively smaller capacities and are also riskier than MSMEs managed by men. On the other hand, Kholifah (Kholifah, 2011) found that the insufficient knowledge, experience, and information networks FMSMEs owned related to the capital mechanism might cause the low level of access to capital perceived by FMSMEs. Consequently, FMSMEs are more likely to face administrative and procedural obstacles in the stages of loan applications, this study proposes the following hypothesis: H3: Gender affects the ease of access to the capital of MSME entrepreneurs, in which MMSMEs perceive a better level of the ease of access to the capital than FMSMEs

Gender, the Ease of Access to Capital, and Patriarchy
Patriarchy rooted in society has rationalized the existence of greater structural control rights for men relative to women. In a society with a high level of patriarchy, women are also more likely to be perceived negatively and treated discriminatively on many aspects of their lives (Sakalli-Ugurlu & Beydogan, 2010). Sakina and Siti (2017) revealed that the existence of patriarchal culture has made women only have little influence-or even no rights at all-in the general areas of society, both economically, socially, politically, and psychologically. The inequality of roles and rights between men and women later became one of the structural obstacles that caused individuals in the community not to have equal access to certain things, including one in terms of business capital (Ardhanari, 2007). This study suggests the potential moderating role of patriarchy on the relationship between gender and the ease of access to the capital of MSMEs. The analysis of the business capital environment situation in Indonesia conducted by Rachmina (2009) shows that the existence of patriarchal culture has limited women's accessibility to capital programs organized by the government. In terms of credit policy, several programs are found to be biassed towards female target groups, such as the Urban Poverty Reduction Program (Program Penanggulangan Kemisikinan Perkotaan-P2KP), Independent Business Performance Program (Karya Usaha Mandiri-KUM), and the Small Farmers and Fishermen Income Improvement Program (Program Peningkatan Pendapatan. . It is due to the general idea that under competitive conditions, the level of potential for success in women-run businesses is relatively lower than that of men. Likewise, the patriarchal environment has made women perceived as more appropriate to act as managers of household affairs, and only men have the right and obligation to work to meet their family needs. Consequently, the number of women who have businesses is found to be relatively lower, while in terms of credit accessibility women entrepreneurs who manage to obtain loans are also in a smaller proportion than men (Rachmina, 2009). This study proposes the following hypothesis: H4: Patriarchy positively moderates the effect of gender on the ease of access to the capital of MSMEs, in which the better level of ease of access to the capital perceived by MMSMEs are getting higher under the conditions of the high level of patriarchy

Gender and the Performance of MSMEs
The issue of MSMEs performance and its antecedents have been investigated extensively in recent years (Ndiaye, Razak, Nagayev, & Ng, 2018). Ardiana, Brahmayanti, and Subaedi (2010) define business performance as the work results that can be achieved by a person or group of people in the organization and is a means in achieving organizational goals. The business performance keeps business actors focus on aspects of business growth and sustainability, which ultimately leads to the fulfillment of business owners' economic goals (Khalid, Maalu, Gathungu, & McCormick, 2016). In various studies, the level of performance of MSMEs can be determined by measuring productivity, developing business units, increasing volume and profitability, or number of sales, employee satisfaction and employee motivation, and employee growth, gross profit and cash flow (Ardiana et al., 2010;Kraus, Rigtering, Hughes, & Hosman, 2012;Rante, 2010). Some of the previous studies have revealed a difference in the level of performance displayed by male and female entrepreneurs of MSMEs. Kim and Sherraden (2014) found that the level of MMSMEs performance was higher than FMSMEs. The superiority performance level displayed by MMSMEs is due to the ownership of business networks with broader characteristics, lower levels of attachment, and the higher potential for resource enhancement compared to the business networks owned by FMSMEs. Nainggolan (2016) and Radipere and Dhliwayo (2014) also found that the level of income of MMSMEs was higher than the income earned by FMSMEs. Compared with MMSMEs, FMSMEs were found to lack in individual personality, which could be the key to success a business, that is, resistance to competitive and challenging conditions and adaptability to risky business environments. This study proposes the following hypothesis: H5: Gender affects the performance of MSMEs, in which MMSMEs display a higher level of performance than FMSMEs.

Gender, Performance of MSMEs, and Patriarchy
This study also suggests the potential moderating role of patriarchy on the relationship between gender and the performance of MSMEs. As Radipere and Dhliwayo (2014) acknowledge, social norms in society can play a role in explaining the differences in performance displayed by men and women of MSMEs. Traditional beliefs that generally still limit women's movement in the area of obligation to fulfil household responsibilities has the potential to make FMSMEs unable to focus on managing the resources they have to achieve performance targets. This condition is in line with the social characteristics that can be found in a patriarchal culture. In a patriarchal society, women are indeed positioned on domestic roles, while men are associated with career achievements, ownership of authority, and leadership dominance (Sakalli-Ugurlu & Beydogan, 2010). Zhu and Kuriyama (2016) found that in countries with environmental characteristics that place a more significant role of women on childcare responsibilities, FMSMEs fight harder to manage their time to develop their businesses. In these countries, women are also assumed to have leadership competencies that are inferior to men, thus creating a gender-based division of work status schemes. This division of work status makes women have fewer opportunities to be involved in their business activities. On the other hand, Klapper and Parker (2011) observed that the social preferences that associate women with household production activities also make women entrepreneurs more likely to utilize the benefits derived from their efforts to meet family needs rather than reinvesting in their business. Meanwhile, the results of the study of Ardhanari (2007) in the micro-business environment in the Surabaya-East Java indicate that the existence of sociocultural factors in the form of family support and the surrounding environment has increased the potential success of businesses run by women micro-entrepreneurs. Based on the above description, this study proposes the following hypothesis: H6: Patriarchy positively moderates the effect of gender on the performance of MSMEs, in which the higher level of performance display by MMSMEs are getting higher under the conditions of the high level of patriarchy

Entrepreneurial Behaviour and Performance of MSMEs
The results of previous studies have revealed a positive relationship between entrepreneurial behaviour and the performance of MSMEs. Rante (2010) proves that entrepreneurial behaviour characteristics like competitiveness, hard work attitude, perseverance, and foresight has a positive influence on the quality of MSMEs performance. In a different study, Kraus et al. (2012) observed that aspects of innovation, the courage to take risks and the ability to manage environmental uncertainties owned by employers were able to keep the MSME business in the Netherlands keep surviving and financially successful in conditions of economic crisis and turbulence. Khalid et al. (2016) also found that the performance of MSMEs was positively influenced by several dimensions of entrepreneurial behaviour such as the high level of risk acceptance, the firm determination for achieving goals and the high level of legitimacy fulfilment. The potential relationship between entrepreneurial behaviour and the performance of MSMEs can be explained using psychological approaches and motivational theories (Delmar, 1996). Psychologically, entrepreneurial behaviour is the actions taken by an entrepreneur based on the ability possessed in order to achieve its business objectives. These actions are believed to arise if only there is a choice that motivates the entrepreneur to act in such a way at a certain intensity and period. These motivations can be in the form of needs for achievement, risk acceptance tendency, tolerance for ambiguity, locus of control, self-efficacy and goal setting or independence and autonomy, reliability and financial security, recognition and status and social-motivational motivations (Shane, Locke, & Collins, 2012;Stephan, Hart, & Drews, 2015). The combination of the ability to act and the motivation that underlies the action-entrepreneurial behaviour -is a significant determinant of the level of performance produced by an entrepreneur (Delmar, 1996). For example, the success of business amid competition conditions will be achieved when the business manager has an excellent ability on organizational resource management based on motivations to accept risk and success and financial security. this study proposes the following hypothesis: H7: Entrepreneurial behaviour of MSME entrepreneurs positively affects the performance of MSMEs

The Ease of Access to Capital and Entrepreneurial Behaviour
As explained in the previous section, entrepreneurial behaviour is the actions taken by an entrepreneur based on the ability and motivation possessed in order to achieve the objectives of his/her business (Delmar, 1996). In other words, entrepreneurial behaviour will arise when there are factors that can motivate someone to act like an entrepreneur. Some researchers (Block & Sandner, 2009;Shane et al., 2012;Stephan et al., 2015;Verheul & Thurik, 2001) argue that the funding/capital aspect is one of the few antecedents to creating entrepreneurial behaviour . Perception of the ease of access to financial capital as a supporting resource for business management is recognized as having an essential role as initial consideration for one's decision to be willing to capture entrepreneurial opportunities in their environment (Shane et al., 2012). Khavul (2010) revealed that difficulties in accessing capital had become the biggest obstacle for MSME actors in developing countries to have the courage in starting and developing their businesses. Block and Sandner (2009) found that the availability of adequate financial resources was able to increase entrepreneurial motivation in Germany to continue to manage their business. On the other hand, Acheampong (2018) also found that a low level of variability in access to funding positively affected someone's decision to start managing a business in Ghana. In a different study, Beck, Lu, and Yang (2015) prove that the ease of access to funding is positively correlated with the decision to become an entrepreneur and an initial investment decision for microenterprises in China. This study proposes the following hypothesis: H8: The ease of access to capital perceived by MSME entrepreneur positively affects entrepreneurial behaviour

The Ease of Access to Capital and Performance of MSMEs
The issue of financing becomes the most significant challenge that is often faced by MSMEs and have a significant impact on its performance (Ndiaye et al., 2018). Adawiyah (2011) found that MSMEs tended to experience difficulties in obtaining capital loans from external parties such as banks or other financial institutions. Hossain, Yoshino, and Taghizadeh-Hesary (2018) observed that the capital environment for MSME in Bangladesh and found that access to funding had a positive effect on the performance of MSMEs. Increasing the level of ease of access to funding, which is marked by the expansion of bank services as a provider of credit, has been proven to increase productivity and business growth. MSMEs that successfully receive loans due to the ease of access are perceived to be more likely to implement a proactive strategy in increasing sales so that, at the very least, they can repay the loans they receive.
In the perspective of Resource-Based Theory, ownership of rare and irreplaceable resources is a necessity for any business in order to obtain a competitive advantage in an industry or particular market scope (Grant, 1999;Kamasak, 2015;Terziovski, 2010). Business capital ownership as a form of financial resources (Acheampong, 2018;Baumane-Vitolina & Cals, 2013), is a fundamental aspect for the company in determining its competitive strategies in order to win opportunities for business success. On the other hand, Okura (2008) states that external funding sources are related to better business growth, where financial flexibility is significantly correlated with business success. In the context of MSMEs, investment from third parties is a source of capital that brings a higher level of business success so that it can be said that when MSMEs have good access to external capital, the productivity and performance of MSMEs will increase. Based on these descriptions, this study proposes the following hypothesis: H9: The ease of access to capital perceived by MSME entrepreneurs positively affects the performance of MSMEs

Data Collection Technique
This study is quantitative research which collects the data through self-administered questionnaires. The questionnaires were distributed (taken) directly to (from) respondents 2 , i.e., MSME entrepreneurs in all districts/municipalities throughout the Province of Bali, that is in Jembrana Regency, Buleleng Regency, Karangasem Regency, Klungkung Regency, Bangli Regency, Badung Regency, Gianyar Regency, Tabanan Regency, and Denpasar City. Of the 300 questionnaires distributed, a total of 298 questionnaires were collected 3 . The next test found that several 97 questionnaires could not be used because there were several items of questions that were not adequately answered or respondents stated that they had never made a credit application to a particular financial institution 4 . Thus, 201 questionnaires are feasible to be tested statistically, and this amount illustrates the value of an adequate response rate (67%).
2 For the purpose of anticipating a non-response bias issue, the return time of the questionnaire is categorized into two groups, namely returns on "Week 1" and "> Week 1". Respondents' answers to each of the latent variables in these two groups were then tested using the non-parametric Mann-Whitney test test. The results show that there were no significant differences in respondents' answers to the Ease of Access to Capital (sig = 0.272), Entrepreneurial Behavior (sig = 0.454) and the MSMEs Performance (sig = 0.967). 3 Two questionnaires were lost at respondent's hands 4 The criterion that the respondent has applied for credit must be fulfilled because it is related to one of the research variables namely "the Ease of Access to Capital". The majority of respondents were male (51.2%) and had the last educational background at the level of elementary/junior high/senior high school (65.7%) with an average age of 41.19 years. Respondents were spread across nine regencies/municipalities in the Bali Province with the most significant proportion being in the Buleleng Regency (51.2%), categorized in the type of business services (15.9%), trading businesses (41.3%) and manufacturing businesses (42.8%), and most have run businesses in the span of 0 -10 years (59.7%). The results of a non-parametric Kruskal-Wallis test did not find any significant differences between respondents' answers to each endogenous variable regarding aspects of educational background, age, business location, type of business and length period of running a business.

Definition and Measurement of Variable
This study operationalizes two exogenous variables (Gender and Patriarchy), and three endogenous variables (Entrepreneurial Behaviour, The Ease of Access to Capital and MSME Performance). In order to improve the validity and reliability of the research construct, all the research variables were measured using research instruments that had been operationalized in previous studies with some minor modifications in term of wording in order to adjust to environmental conditions and research subjects.
Gender is a dummy variable measured using a nominal scale with "0" means "Male," and "1" means "Female." The variable of Ease of Access to Capital refers to the level of ease of obtaining information, services, and approval of requests for funds (capital) submitted by MSME entrepreneurs. Items were adapted from a questionnaire on the Access to Finance of Enterprises Survey (SAFE) developed by the European Central Bank (European Central Bank, 2017). This variable measurement is done by asking the respondent to indicate their level of agreement using a 5-point Likert-type scale, ranging from "Never" to "Very Often" on five statements related to conditions experienced when applying for credit/loan funds to the bank, Village Credit Institutions (Lembaga Perkreditan Desa-LPD) or other financial institutions.
Entrepreneurial Behaviour is defined as actions carried out by business managers based on the desire to develop their business armed with leadership and managerial capacities to achieve business goals amid the existing competitive conditions (Gray, 2002). This variable is measured by asking respondents to indicate their agreement using a 5-point Likert-type scale, ranging from "Strongly Disagree" to "Strongly Agree" on four statements related to entrepreneurial characteristics applied by respondents in managing the organization. The items used were adapted from the instruments developed by Rante (2010).
Patriarchy refers to the universal values that apply and is accepted in society for the existence of structural control that is more dominantly owned by a male-relative to female-over political, legal, economic and religious aspects (Sakalli-Ugurlu & Beydogan, 2010). This variable measurement is done by asking the respondent to indicate their level of agreement using a 5-point Likert-type scale, ranging from "Never" to "Very Often" on two statements in the form of assumptions/perceptions that are often felt/heard in the respondent's daily environment related to the patriarchy. The items are: "Female should prioritize marriage and family matters compared to career and decision to work outside the home" and "If a female works, then the condition of the household and her family will be threatened." MSMEs performance is defined as the work results that can be achieved by a person or group of people in the organization and is a means of determining in a process to achieve organizational goals (Ardiana et al., 2010). The instrument used consists of three items that require respondents to rate the level in which (1) the number of sales obtained always increases every year; (2) employees are always satisfied with the salary and bonus received, and (3) employees have high motivation in work. Assessment by respondents of the statement items was carried out using a 5-point Likerttype scale, ranging from "Strongly Disagree" to "Strongly Agree."

Hypotheses Testing Technique
In order to test the proposed hypotheses, this study conducts a structural equation modelling techniques (SEM) with the partial least square (PLS) approach. In general, the use of the PLS aims to maximize the variance explained by the variables in the model by using the amount of R-Square (R 2 ) as a measure of quality (goodness-of-fit) of the research model formulated (Chin, Marcolin, & Newsted, 2003). PLS has advantages in terms of flexibility and tolerance for a relatively small number of samples and the problem of multicollinearity between independent variables and does not require the fulfilment of the assumption of normal data distribution. PLS simultaneously able to test the measurement model (the relationship between latent variables (constructs) with each indicator) as well as the structural model (the relationship between constructs) so that the validity of the entire construct can be adequately assessed (Chenhall, 2005). This study uses the SmartPLS 3 Professional Version application to test hypotheses.

Preliminary Data Analysis
Before designing and testing the overall model, a preliminary data analysis is carried out in order to ensure the quality of the proposed model. Table 1 displays the results of descriptive statistical analysis for each variable which includes the mean, standard deviation (SD) and the minimum and maximum scores of the respondent's answers (both theoretical and actual). Based on Table 1, the high mean score and low standard deviation generated for the variable of entrepreneurial behaviour (mean = 4.3184; SD = 0.35954) and the performance of MSMEs (mean = 3.8905; SD = 0.45198) indicate that the MSME entrepreneurs in the Bali Province tended to have a high level of entrepreneurial behaviour and a relatively satisfactory level of performance. Meanwhile, the ease of access to capital perceived is also quite good (mean = 3.7294) with perceptions of the level of patriarchy felt to be at a reasonably high level (mean = 3.3184). In addition, a bivariate correlation analysis is also performed based on the average scores of the variables. As presented in Table 2, the results of the correlation analysis provide initial support for the hypotheses that predict a negative relationship between female gender and the level of entrepreneurial behaviour (H1) (α = -0,144; p-value < 0,05), the ease of access to capital (H3) (α = -0,222; p-value < 0,01) and the performance of MSMEs (H5) (α = -0,145; p-value < 0,05). Likewise, the hypothesis that predicts the moderating role of patriarchy on the relationship between gender and the ease of access to capital (H4) (α = -0,178; p-value < 0,05) and the positive relationship between entrepreneurial behaviour and MSMEs performance (H7) (α = 0,471; p-value < 0,01) also appears to be supported by observing the related sign and significance value of the Pearson Correlation coefficient in Table 2.   Diagonal element: square root of AVE; off-diagonal: correlation between constructs.

Measurement Model Analysis
An assessment of the fit indices yields the Standardized Root Mean Residual (SRMR) index value of 0.08, and the Normed Fit Index (NFI) of 0.719. Therefore, the model can be declared to have an adequate level of feasibility. The criteria for convergence validity have also been fulfilled by considering the amount of loading factor of specific construct indicators that have greater value (> 0.5) compared to other indicators in the construct that is not associated with it (the cross-loading value). Likewise, as presented in Table 3, the score of average variance extracted (AVE) (minimum 0.655) and Cronbach's alpha (minimum 0.629), rho_A (minimum 0.646) and composite reliability (minimum 0.842) have met the minimum criteria required. Meanwhile, the results of the comparison between the square root value of AVE (the value showed on the diagonal element of Table 4) and the correlation value between the two constructs (the value showed on the 0ffdiagonal element of Table 4) also support the fulfilment of the discriminant validity of the measurement model. Hair, Black, Babin, and Anderson (2010) contend that the square root value of AVE must have a higher value than the correlation value between the two constructs for the measurement model to be met the discriminant validity requirement.

Structural Model Analysis
The value of adjusted R 2 (R-squared) of the dependent (endogenous) construct can then be used as a qualification criterion in evaluating the structural models that have been developed. As shown in Table 3, for the construct of Ease of Access to Capital (EASE), the value of R 2 is 9.4%, the construct of MSMEs Performance (PERF) is 12.7%, and the construct of Entrepreneurship Behaviour (BHVR) is 0.6%. The results of the R 2 amount explain that the proposed research model can explain the variation in the EASE construct of 9.4%, the PERF construct of 12.7% and the BHVR construct of 0.6%; while the remainder is explained by other variables outside the proposed model.

The Effects of Gender on Entrepreneurial Behaviour
The results of PLS analysis (Table 5) indicate that the first hypothesis (H1) in this study cannot be empirically supported by considering the path coefficient value that is negative yet not significant (γ1 = -0.082; p-value = 0.251). This is in line with the results of non-parametric Mann-Whitney test (Table 6 Panel A) for the entrepreneurial behaviour variable which shows a higher mean rank value (MR) in the male group (MR = 107.90; n = 103) compared to the female group (MR = 93.75; n = 98), yet not significant (p-value = 0.082). Thus, it can be stated that gender does not affect entrepreneurial behaviour, in which there is no difference in the level of entrepreneurial behaviour displayed by male and female entrepreneurs of MSMEs in the Province of Bali. These results are not consistent with Kim and Sherraden (2014), Selamat, Razak, Gapor, and Sanusi (2011) and Sharma, Dua, and Hatwal (2012), but consistent with Suwarno (2015).

The Moderating Role of Patriarchy on the Relationship between Gender and Entrepreneurial Behaviour
The results of PLS analysis (Table 5) indicate that the second hypothesis (H2) in this study cannot be empirically supported by considering the path coefficient value that is not significant (γ2 = -0.028; p-value = 0.709). This finding is also supported by the results of the Moderated Regression Analysis (MRA) test (Table 7 panel A) which indicates that there is no significant two-way interaction between gender and patriarchy in influencing entrepreneurial behaviour (p-value = 0.726). Thus, it can be stated that patriarchy does not play a moderating role in the relationship between gender and entrepreneurial behaviour, in which differences in perceived levels of patriarchy does not significantly widen the distance between the level of entrepreneurial behaviour displayed by MMSME and FMSME in Bali Province. This result is not consistent with the research of Ardhanari (2007) and Kholifah (2011).

The Effects of Gender on the Ease of Access to Capital
Meanwhile, the third hypothesis (H3), which predicts a significant relationship between gender and the ease of access to capital can be stated to be empirically supported (p-value = 0,000). Negative path coefficient value (γ3 = -0.224) indicate that FMSMEs tends to perceive lower levels of the ease of access to capital than MMSMEs do. This is also supported by the results of non-parametric Mann-Whitney test (Table 6 Panel B) which shows a higher mean rank value (MR) in the male group (MR = 111.42; n = 103) compared to the female group (MR = 90.05; n = 98) for the ease of access to capital variable. The difference in the mean rank of these two groups was significant at the 1% significance level (p-value = 0.008). Thus, it can be stated that gender influences the ease of access to capital, in which there are differences in the level of ease of access to capital perceived by male and female entrepreneurs of MSMEs in the Province of Bali. MMSMEs were found to perceive a better level of the ease of access to capital than FMSMEs. This result is consistent with the study of Ardhanari (2007), Kholifah (2011) and Nainggolan (2016).

The Moderating Role of Patriarchy on the Relationship between Gender and the Ease of Access to Capital
A prediction of the moderating role of patriarchy on the relationship between gender and the ease of access to capital as proposed in the fourth hypothesis (H4) can be stated to be empirically supported (γ4 = 0.141; p-value = 0.047). This finding is also supported by the results of the Moderated Regression Analysis (MRA) test (Table 7 panel B) which indicates a significant two-way interaction between gender and patriarchy in influencing the ease of access to capital at a significance level of 5% (p-value = 0.020). In addition, the results of the MRA indicate that the interaction is positive (b3 = 0.252), which means that patriarchy strengthens the influence of gender on the ease of access to capital. In other words, the higher the level of patriarchy that is felt, the lower the ease of access to capital felt by FMSMEs and, simultaneously, the higher the ease of access to capital perceived by MMSMEs 5 .

The Effects of Gender on MSMEs Performance
The results of the PLS analysis (Table 5) show that the fifth hypothesis (H5) in this study can be stated to be empirically supported by considering the value of the path coefficient which is negative and significant at the 1% significance level (γ5 = -0.213; p-value = 0.000). Negative path coefficient values indicate that FMSMEs tends to perceive lower levels of achieved performance than MMSMEs. This finding is also supported by the results of non-parametric Mann-Whitney test (Table 6 Panel C) which shows a higher mean rank (MR) value in the male group (MR = 109.35; n = 103) compared to the female group (MR = 92.22; n = 98) for the MSMEs performance variable. T The difference in the mean rank of these two groups was significant at the 5% significance level (p-value = 0.036). Thus, it can be stated that gender affects the performance of MSMEs, in which there are differences in the level of performance displayed by male and female entrepreneurs of MSMEs in the Province of Bali. MMSMEs were found to show a better level of performance than F MSMEs. This result is consistent with the research of Minniti and Nardone (2007) and Nainggolan (2016).

The Moderating Role of Patriarchy on the Relationship between Gender and the MSMEs Performance
The sixth hypothesis (H6) that predicts the moderating role of patriarchy on the relationship between gender and the performance of MSMEs can be stated to be empirically unsupported (γ6 = -0.035; p-value = 0.607). This finding is also supported by the results of the Moderated Regression Analysis (MRA) test (Table 7 panel C) which indicates that there is no significant twoway interaction between gender and patriarchy in influencing MSME performance (p-value = 0.593). Thus, it can be stated that patriarchy does not play a moderating role in the relationship between gender and MSME performance, with differences in perceived levels of patriarchy does not significantly widen the distance between the different levels of performance displayed by MMSMEs and FMSMEs in Bali Province. These results are not consistent with the study of Ardhanari (2007) as well as Zhu and Kuriyama (2016). 5 An additional analysis was carried out by comparing the magnitude of differences in the scores of ease of access to capital perceived by male and female entrepreneurs of MSME on the category of low and high levels of patriarchy. As done by Sakalli-Ugurlu and Beydogan (2010), this categorization is based on the median score of the patriarchy (3.25), which is a low category for the respondent group with the value of the patriarchy score below 3.25 and vice versa. The results showed that the range of differences in the scores of of ease of access to capital perceived by MMSMEs and FMSMEs was found to be greater in the category of high level of patriarchy (0.3403) than in low level of patriarchy (0.3147).

The Effects of Entrepreneurial Behaviour on MSMEs Performance
Regarding the seventh hypothesis (H7), which predicts a positive relationship between entrepreneurial behaviour and MSMEs performance, the results of PLS analysis, as presented in Table 5, indicate some empirical support. The positive (γ7 = 0.304) and significant (p-value = 0,000) path coefficient values indicate that the higher the level of entrepreneurial behaviour possessed, the higher the level of performance that can be achieved by MSME entrepreneurs. Thus it can be stated that entrepreneurial behaviour proved to have a positive effect on the performance of MSMEs in the Province of Bali. This result is consistent with the research of Sumantri et al. (2013), Rante (2010), Kraus et al. (2012) and Khalid et al. (2016).

The Effects of the Ease of Access to Capital on Entrepreneurial Behaviour
The results of PLS analysis (Table 6) show that the eighth hypothesis (H8) in this study can be stated to be empirically supported by considering the path coefficient value which is positive and significant at the 1% significance level (γ8 = 0.281; p-value = 0,000). These results indicate that the higher the perception of the ease of access to capital felt, the higher the level of entrepreneurial behaviour displayed by MSME entrepreneurs. Thus it can be stated that the ease of access to capital has a positive effect on entrepreneurial behaviour of MSME entrepreneurs in the Province of Bali. This result is consistent with the research of Block and Sandner (2009), Beck et al. (2015) and Acheampong (2018).

The Effects of the Ease of Access to Capital on MSMEs Performance
Meanwhile, the ninth hypothesis (H9), which predicts a significant relationship between the ease of access to capital and the performance of MSMEs can be stated to be empirically unsupported (pvalue = 0.098). Thus it can be stated that the ease of access to capital has proven to not affect the performance of MSMEs in the Province of Bali. This result is not consistent with the research of Adawiyah (2011) and Okura (2008).

Conclusion
This study examines the relationships between gender, patriarchy, entrepreneurial behaviour, the ease of access to capital, and the performance of micro, small, and medium enterprises (MSMEs). By using data from 201 micro, small and medium enterprises (MSMEs) throughout the province of Bali, the results indicate that (1) women entrepreneur of MSMEs perceive lower levels of the ease of access to capital compared to male entrepreneur of MSMEs; (2) patriarchy positively moderates the relationship between gender and the level of the ease of access to capital; (3) female entrepreneur of MSMEs display lower levels of performance than male entrepreneur of MSMEs; (4) entrepreneurial behaviour has a positive effect on the performance of MSMEs; and (5) the ease of access to capital has a positive effect on entrepreneurial behaviour . Unfortunately, the results cannot provide support for the research hypothesis related to (1) the influence of gender on entrepreneurial behaviour; (2) the positive moderating role of patriarchy on the relationship between gender and entrepreneurial behaviour; (3) the positive moderation role of patriarchy on the relationship between gender and the performance of MSMEs; and (4) a positive relationship between the ease of access to capital and the performance of MSMEs. The findings of this study provide a practical overview of the capital environment for MSMEs in the Province of Bali in particular, and Indonesia in general. The relatively lower perception of the ease of access to capital felt by the female entrepreneur of MSMEs indicates that a conducive business climate is still not able to be created when considering aspects of gender equality. Policies that provide more opportunities and equal treatment for women MSMEs in Indonesia, especially related to business funding, still need to be increased in quantity and quality. Likewise, taking into account that entrepreneurial behaviour is proven to have a significant influence on the performance of MSMEs, educational institutions need to provide more portion of entrepreneurial materials to the structure of their curriculum. In addition, the government, through relevant institutions, needs to ensure the creation of entrepreneurial business opportunities and funding grant programs that can stimulate entrepreneurial motivations for the younger generation in Indonesia.
This research is not without its limitations. First, the use of survey methods through questionnaires is so vulnerable to response biases (Jogiyanto, 2011). However, this study has attempted to minimize the possible bias through direct distribution and collection of instruments and conducting a Kruskal-Wallis test with adequate results to be continued at the hypothesis testing stage. Future studies can apply mixed methods research strategies by conducting interviews to confirm the answers given by respondents to the questionnaire so that they can obtain more comprehensive results and conclusions. Second, this study has not considered the differences in capacity and needs of MSMEs in each business category, namely micro, small, and medium enterprises. Future studies can examine the possibility of differences in the relationship between variables on the same topic in each of the MSME business categories.