Female Executives in the Service Sector: The Case of Small and Medium-Sized Enterprises in Spain

: Small and medium-sized enterprises (SMEs) and women are important agents for economic and social development. This study attempts to understand how the economic crisis has a ﬀ ected the behavior of SMEs with high female participation in the decision-making process. To that end, the quantitative research focused on a sample of 425 small and medium-sized enterprises in Spain relying on a questionnaire-based survey. The results of the empirical study o ﬀ er new evidence on the e ﬀ ects of the recession on decision-making female-dominated bodies in organizations and illustrate the importance of corporate governance issues. In addition, the research provides greater knowledge of female entrepreneurs in a country severely a ﬀ ected by the global economic crisis. Several recommendations are proposed to provide a theoretical basis for governments to reduce glass ceiling and promote women’s empowerment policies. Author Contributions: Conceptualization, L.E.-S. and A.F.G.-C.; methodology, L.E.-S.; writing—original draft, L.E.-S. and A.F.G.-C..; writing—review and editing, L.E.-S. and A.F.G.-C.; visualization, A.F.G.-C..; project administration, L.E.-S. and A.F.G.-C.; funding acquisition, L.E.-S. and


Introduction
Gender diversity is an important force in achieving sustainable development. Moreover, sustainable development must include the promotion of women's empowerment and gender equality [1]. Despite the increasing attention that corporate governance diversity has received in recent years, the majority of studies address managing to grow at successful organizations [2,3]. Nevertheless "there is also an opportunity to significantly augment researchers' understanding of the period following a crisis" [4] (p. 378). Although new strategies and policies are required, there is little research on corporate governance structures and what allows firms to successfully go out of a financial crisis [4][5][6]. In addition, much research on corporate governance management has addressed large and listed firms [7], and the study of small and medium-sized enterprises (SMEs) has focused principally on developed economies [8,9]. On the other hand, small firms have unique specificities that differentiate them from larger firms [10]. As Gilmore et al. [10] (p. 7) state, these features can be determined by "the inherent characteristics and behaviors of the entrepreneur or owner/manager; and they may be determined by the inherent size and stage of development of the enterprise". Regarding industry, Hisrich and Brush [11] found that the probability that a woman controls a company is greater if this firm belongs to the retail or services sector than to others.
Another stream of research is that addressing companies set up and run by women. Since the eighties, Spain has undergone significant political, social and economic changes. The number of women managers and professionals grew by 81 percent between 1984 and 1990, and during this period, the number of women business owners increased by 57.8 percent [12]. Before the financial crisis, Blasco-Burriel et al. [13] analyzed female entrepreneurs' perceptions of the reasons for there being fewer women entrepreneurs in the Aragón region in Spain and found that the explanation lied in family responsibilities. For these women, self-employment was relatively unattractive for women with family responsibilities. However, there is still an under-representation of women in decision-making positions, which implies a need for academic studies on the conditions that favor the female presence. Moreover, Spain is below the EU-28 average at the level of women on boards (23.7% versus 26.7%), as well as at the decision-making bodies (15.8% versus 17.6%), according to the 2019 Gender Statistics Database of the EU European Institute for Gender Equality [14].
Additionally, there is a scarcity in the studies about how women run enterprises and perform as managers and small business owners [15][16][17]. SMEs are often the hardest affected companies of a prolonged economic crisis [18][19][20]. These firms are more vulnerable because they lack the resources and management mechanisms to deal with unexpected events [21]. According to Pricop [22] (p. 106), "the strategic management theory should focus more on the casuistry of small and medium-sized companies which have an overwhelming percentage in the number of organizations from an economy". Little work has been carried out to explore strategic management focused on SMEs during difficult times. In the Organisation for Economic Co-operation and Development (OECD) economies, SMEs account for over 95% of firms (99% in Spain), two-thirds of total EU-28 employment (66.4 %), figure that in Spain reaches above 70% [23].
In this article, we explore the impact of the 2008 global financial crisis on the presence of female executives in the Spanish context, given that "the relative inattention to firms in crisis is unfortunate" [4] (p. 377). The study draws on a sample of Spanish SMEs to advance our understanding of the characteristics of companies run by women. The research focuses on the service sector, where there are more women owner-managers than in other industries [13]. In addition, gender diversity can have a higher competitive advantage in the service sector than in the manufacturing industry [24] (p. 1470) due to factors such as greater value of market insight and higher interaction among human resources and costumers in services firms than in manufacturing companies [25].
To this end, we formulated a questionnaire in 2008 about strategic business decisions, and now we try to predict the consequences of these actions on the position of women executives in 2013. We address three related research questions: (1) Based on strategic choices made in 2008 or previous years, is it possible to predict whether the CEO will be a woman in 2013? (2) Does the implementation of measures of conciliation of labor and family life in 2008 determine what firms will be women-run businesses in the future? (3) Will the position of the company respect to the competition influence the future of women executives?
This study makes several significant contributions to the growing literature on female executives in several ways. First, this article provides new evidence about the consequences of strategic decisions made by Spanish SMEs at the most critical moment since the 2008 global financial crisis, on female executives. Second, unlike previous studies in the strategy literature related to top management teams based on the influence on the entire team rather than on one individual person, we focus on one single person, the CEO, given the weight that this figure has in small and medium enterprises. While previous studies have measured the level of firm innovation through exploring corporate boards as decision-making groups (e.g., [26,27]), this article focuses on female CEOs as one individual person. Third, the study fills a significant gap in strategic decisions on female CEOs participation in Spanish SMEs. The current study aims to extend this literature by analyzing previous strategic decisions on the gender of future leaders. To this end, we have adopted a multi-theoretical framework [28,29], drawing from insights provided by agency, legitimacy, stakeholder, institutional and signaling theories in conducting our analysis.
The results of the study show that businesses run by women are positively and significantly related to prior adoption of measures of conciliation of labor and family life, pursuing a strategy based on expanding the geographical coverage and being firms more innovative than companies from competition.
The remainder of this article is organized as follows. Section two contains a discussion of the literature relevant about gender diversity in corporate governance and the importance of diversity in organizational innovation. Section three develops the research hypotheses and describes the research methods. Section four presents and analyses the empirical results, and finally, section five presents the discussion and conclusion.

Gender Diversity and Corporate Governance
There are various approaches from which corporate governance issues have been addressed. A substantial literature has focused on agency theory [30][31][32][33] and on how to reduce the agency costs. This theory suggests that when ownership and control are unified agency costs are reduced [31]. As Li [34,35] indicated, mutual monitoring between top executives is an important substitute for other governance mechanisms since it can mitigate agency problems. The potentially self-interested CEO can be monitored and constrained by the number two executive in a firm, who is more likely to be female [35]. Nevertheless, in the context of SMEs, smaller agency costs are expected due to the minor separation of ownership and control and the minor conflicts in the management relationship. In these firms, the CEO's role is essential. According to Martín-Ugedo et al. [36] (p. 3), "the CEO may be more important than the board in the running of a company because he/she deals with the day-to-day operations of the firm. In addition, in SMEs the board is less likely to constrain the CEO". Female participation studies in the decision-making process have long been concentrated on listed companies, and there are fewer studies based on SMEs. It is important to note that in the corporate governance studies, an extensive research stream is focusing on the topic of women on boards of directors [37][38][39].
Moreover, from a broader perspective, other approaches such as legitimacy, stakeholder, signaling and institutional theories have been used in the studies about corporate governance and its connections with the role of women in company decision-making. While agency theory mainly focuses on shareholders rather than on a broader spectrum of stakeholders, "social and political theories", such as those mentioned above, also consider a broader set of agents. Under these perspectives, the gradual introduction of women in decision-making positions and its relations with corporate governance issues can be addressed. Stakeholder theory [40] suggests that organizations have to meet multiple expectations of their various stakeholder groups, rather than only those of their shareholders, going beyond simple economic or financial performance. Society and governments are demanding higher diversity and gender empowerment in business, as well as in other areas [41].
Signaling and legitimacy theories have also been used in the context of gender and corporate governance. Signaling theory [42][43][44] suggests that companies will try to reduce the information asymmetries by sending signals to interested parties. Moreover, legitimacy theory [45][46][47] is based on a "social contract" between the organization and society at large so that the survival of the organization is determined by both market forces and society's expectations. The gender profile of the company will act as a signal for the market and potential executives' candidates. Additionally, gender inclusion and ceiling glass reduction are social tendencies that companies should take into account not only because of the potential direct effects on their performance but also due to social and legitimacy requirements of the society in which they operate. In the same way, stakeholders have an influence in both social norms generally accepted in the organization and institutional practices. Thereby, according to institutional theory [48,49], organizations adopt and adhere to standard institutional practices in order to enhance or protect their legitimacy by conforming to the expectations of institutions and stakeholders.
Previous research on the diversity in the boardroom has often focused on tokenism theories [64][65][66], and numerous studies have addressed the barriers facing women in the upper echelons of organizations [65,67,68]. Concerning compensation incentives as governance, Core and Guay [69] develop a model for the optimal level of equity incentives for CEOs. Some authors suggest a positive association between external industry pay gap and firm performance, firm risk, the riskiness of firm investment and financial policies, offering evidence of the external interactions between CEOs in the industry tournament [70]. Other authors highlight the importance of the level of competitiveness of the industry when studying the impact of corporate governance on the efficiency of the firm. For example, Giroud and Mueller [71] find that firms in non-competitive industries benefit more from good governance than firms in competitive industries do.
Several studies have examined gender differences in executive pay among male and female managers. Some of them found that men executives are compensated at higher levels than women [72,73]. Conversely, Bugeja, Matolcsy, and Spiropoulos [74] found no difference in compensation for female CEOs, and others found differences in its composition [75]. According to some authors, once the women have been able to climb the corporate ladder to the top, female executives face no gender bias in remuneration. On the contrary, before the economic crisis of 2008, Lee and James [76] suggested that the reactions of the investors were more unfavorable to the appointment of female than male CEOs.
On the other hand, the level of influence of the executive director on the firm's strategy is another critical issue. Lee and James [76] (p. 228) maintained that notwithstanding a pipeline of female executives is almost on par with their male peers, the distribution of women in top management and executive positions is strongly biased in favor of men. Moreover, despite the 2008 financial-economic crisis, research suggests cautious optimism about the future of women managers [77].

Women's Participation in Organizational and Innovative Strategies
Another line of research has focused on the relation between the impact on the strategic choice on innovation fields of firms and top management teams' diversity [78]. Most of these studies focus on the entire team rather than to one individual person [78][79][80]. Muller-Kahle and Schiehll [81] (p. 667) also mentioned that "the drivers or determinants of female CEOs' power in organizations have been understudied", and the relation of women directors and female executive has been largely neglected in existing studies to date [82] (p. 452). However, as pointed out by Wang and Kelan [82] (p. 449) "a particular area of interest within this stream of the literature has been women's participation in organizational decision making as a result of holding leadership positions". Other research highlights the importance of gender diversity in organizational innovation [83,84]. Nakagawa [85] found in a study of Japanese women managers a positive relationship between women's managerial participation and firm performance, thus gender diversity could provide a source of competitive advantage.
Torchia et al. [84] analyzed the contribution of female directors to firm innovation by strategic board tasks and found that the level of firm organizational innovation is higher when there is a critical mass of women (at least three). Na and Shin [86] found a positive relationship between female top management and marketing innovation as well as between female ownership and several innovation measures in emerging countries. In addition, the literature on corporate governance suggests that the effects of ownership and governance variables could impact innovation in SMEs differently than in large companies [87]. Moreover, for given R&D expenditure, Chen et al. [88] found greater innovative success when there is a greater representation of women in the board of directors.

Reconciliation of Work and Family Life
Schwartz [89] offers several reasons why talented women make sound business sense. Regarding female entrepreneurs, some authors find that, from an integrated perspective, to be an entrepreneur is compatible with being a mother, wife or daughter [90,91]. Some women become entrepreneurs because, in this way, they have flexibility, facilitating the balance between work and family [91,92]. Another reason that leads to an increase in the number of female entrepreneurial initiatives is a reaction against persistent discrimination in the labor market [93]. Regarding differences between men and women, for some authors, women seek a balance between work and family life, while men are more motivated by the creation of wealth or by economic progress [94][95][96][97]. According to Clevenger and Singh [98], the conflict between work and family could be one of the causes of the glass ceiling.
In Spain, the results of a survey carried out in the early nineties showed that the obstacles that faced women executives to advance their professional careers were, "discrimination, the male culture of organizations and male communication networks, family ties, the timetable system and self-imposed limitations" [12] (p. 199). As Muller-Kahle and Schiehll [81] (p. 668) note, "Some studies addressing the glass cliff concept suggest that, given the extremely demanding and challenging corporate environments in which women generally work, they tend to accept CEO and other leadership positions in companies with high potential for failure". On the contrary, other studies suggest that women are more risk-averse [75,99,100], which implies a preference for strategic decisions with less assumed risk. Therefore, women are supposed to be more conservative as entrepreneurs than men because of their greater aversion [101][102][103][104], and they will be more attracted to those companies in which non-related diversification strategies are run. Geographical diversity offers a reduction of the global risk of the company. On the contrary, policies of growth in the same market/product suppose an increase in the global risk of the company.
Moreover, females may be particularly sensitive to certain social responsible organizational practices [39], such as those related to the reconciliation of work and family life. Therefore, on the one hand, those companies with sensitivity towards conciliation will value feminine profile features. On the other hand, signaling theory [42][43][44] offers support to the higher presence of women in executive positions on those companies. Potential female candidates will positively value information about the responsible social preference of the firm when they consider applying for an executive position. As Srinidhi et al. [105] point out "in general, female participation on boards is seen to be favourable for several reasons, such as females (relative to males) are more sensitive to charitable and community matters, have a wider educational and work experience background and communicate in a more participatory manner which often encourages a broader perspective and inclusion of stakeholder needs". Reconciliation opportunities as part of the total compensation model may act as a signaling element, given that these measures are highly assessed by women.
Stakeholder theory suggests that a company is required to respond to a multiplicity of stakeholders [40]. Additionally, according to the legitimacy theory, companies should take into account the trends and demands of society and frame their activities within the social changes. The trend towards the introduction of women's participation in decision-making positions is justified both by the social requirements of female representation and for the possible improvements in the company's performance. Therefore, it is expected that the decisions of the organization are consistent with gender diversity trends. As Kartadjumena and Rodgers [106] argue, companies can improve their management decisions when the interests of different stakeholders are taken into account, in addition to shareholders. For Daily, Certo and Dalton [107], women could better understand the needs of female customers. James [108] explores the advantages of learning and innovation resulting from work-life balance provision in knowledge-intensive companies, as part of a mutual gains research agenda. James [108] (p. 288) found that "by making available the kinds of work-life balance arrangements self-reported by workers as offering meaningful reductions in gendered everyday work-life conflicts, employers can also enhance the kinds of learning and innovation processes that are widely recognized as fundamental to firms' long-term sustainable competitive advantage".

Study Design and Hypotheses Development
This study investigates the effects of several variables related to the business strategy followed in 2008, the year hardest hit by the current economic crisis and its impact on the future incorporation of women in management five years later. The objective of this study is to better understand the influence of several strategic decisions in a representative sample of small and medium Spanish enterprises at the most critical moment of the current economic crisis in the future presence of women leaders.
Potential hypotheses are categorized into different groups based on the structure of a questionnaire. In order to test the validity of the strategic decisions, we examine three aspects, past actions, current behavior and practices, and future prospects of SMEs in Spain. Hence, we want to test whether the influence of strategic decisions positively affects the future of female executives in these firms. To test our hypotheses, we developed a questionnaire with 19 items (the items are set out in Table 1)  Responses to the questionnaire were used to construct the variables. The items were developed based on previous dimensions. A seven-scale instrument was used to measure the position of the company compared to the competition, with one being strongly much lower than the competition and seven being strongly much higher than the competition. Moreover, we obtain information from secondary sources for the post-crisis variables.

D3
Does the firm carry out advertising? D4 Does the firm apply any measure of reconciliation of work and family life? What is the position of the firm compared to the competition on the following questions? (with 1 being strongly much lower than the competition and 7 being strongly much higher than the competition) D5 Introducing innovations? D6 Growing in recent years? D7 Offering quality of services? D8 Setting competitive prices? D9 Having reputation and brand recognition? D10 Setting wages? D11 Regarding the degree of employee satisfaction? D12 Providing training to employees?

Questions -Dimension 3: FUTURE STRATEGY
Questions formulated to companies that have affirmed they will expand their business in the coming years How does the firm plan to expand its activity in the coming years? D13 Expanding the size of the current business? D14 Expanding the geographical coverage? D15 Offering products or services related to existing? D16 Offering other products or services not related to existing?

Control variables:
Firm Age Firm Size Number of employees Number of Owners Therefore, we formulate the following hypotheses: (I) Hypotheses relating to pre-crisis strategies: Hypothesis Ia: There is a positive relationship between the firms that have introduced innovative processes in the pre-crisis period and the presence of Female CEO five years later. Hypothesis Ib: There is a positive relationship between the firms that have created new products or improved service in the pre-crisis period and the presence of Female CEO five years later. (II) Hypotheses relating to strategies during the crisis: Hypothesis IIa: Firms with a strong competitive position on strategic issues are more likely to be run by women five years later. Hypothesis IIb: Firms implementing measures to balance working life and family are more likely to be run by women five years later. (III) Hypotheses relating to future strategies: Hypothesis IIIa: Firms having a future growth strategy are more likely to be run by women five years later.
For each hypothesis, there is no single theory but a combination of those mentioned in the theoretical framework. Different authors suggest that the use of more than one approach helps to obtain a greater understanding of the practice [109,110]. As noted by Gray et al. [111] theories should be complementary to each other.

Data and Methodology
This study builds on a representative sample of Spanish small and medium-sized enterprises in a central-eastern province of Spain (Teruel). This data set relies on the SABI (Iberian Balance Sheets Analysis Systems) database in 2008 and in 2015 (the latest information provided is for January 2013). Besides, information about strategic business decisions used in this analysis comes from a survey conducted in 2008, the year in which the economic downturn manifested critically.
The initial data sample consists of all non-financial firms belonging to the services sector in 2008, a total of 1,652 enterprises. Throughout this year, a questionnaire was prepared and sent by post and by electronic mail to all the firms, potential respondents, together with a stamped addressed return envelope in the case of the traditional post mail. When the letters were returned or not answered, we interviewed the possible respondent by phone. At the beginning of the survey, to avoid non-response or self-selection bias, the interviewers placed particular emphasis on the anonymity of the responses and informed that no individualized or confidential data on the results would be distributed. In this way, we tried to avoid the possible non-response of those firms that fear that their information may reach the hands of potential competitors or that it may be misused. The sample necessary for the study, with a confidence level of 95 percent and a low mistake to +-4 for 100, is 441 companies. The final response of 522 represents about a 31.6 per cent response rate, more than usual in this kind of surveys. This supposes an estimation error of 3.55 percent for the mentioned level of confidence. Definitively, it is possible to indicate that the sample represents a great set of companies of the population as a whole, which allows the results to be generalized, although "leading thinkers and methodologists from both the postpositivist and constructivist schools have long recognized that generalizations can never be made with certainty" [112] (p. 1457). According to Polit and Beck [112] no method, qualitative or quantitative, is perfect when it comes to generalizing the results.
In this section, we are trying to build up a model to predict female CEO presence on firms in 2013, using data on strategic decisions obtained in 2008. We are searching for associations between the firms running by women and different items in the survey and other control variables.
The dependent variable is dichotomous; the two categories of CEOs, are male and female.
Since we are interested in the odds of female CEO, the reference category is men. Therefore, women were coded as one and men as zero. Based on the research questions and because our dependent variables are dichotomous, such a female CEO versus male CEO, logistic regression seemed to be the best method of analysis, so our approach draws from logistic regression model to test our hypotheses.
The logistic regression model employs the binomial probability theory. There are two values to predict: that probability (P) is one rather than zero, i.e., the female CEOs belong to one firm rather than the others. A model is created that includes all predictor variables that are useful in predicting the response variable. Logistic regression calculates the probability of success over the likelihood of failure. Since the research question is whether strategic decision variables would affect the odds of CEO differently for men and women, the statistical method employed was binary logistic regression. The general form of a logistic regression equation from which the probability of Y is given by: In which P(Y) is the probability of Y occurring, e is the base of natural logarithms, and β are the coefficients. A value close to zero means that Y is very unlikely to have occurred, and a value to one implies that Y is very likely to have occurred. In our study, the logistic model predicts the probability of a woman being CEO for the ith firm. Therefore, for a given firm, Y will be either zero if the outcome did not occur or one if the outcome did occur. On the other hand, P(Y) will be a value between zero and one, where zero is no chance that the CEO being women and one means that certainly the CEO will be a woman in 2013.
Following the proposed hypotheses, the predictor variables considered in this study are represented by the answers to the questions that were asked to interviewees. The model includes the predictor variables listed in Table 1. We included three control variables that are used in prior studies: firm age, firm size measured by the number of employees, and the number of owners. As stated by Ali [113] (p. 8), "compared to old firms, new firms with less formalized policies and practices may be better positioned to capitalize on the benefits of gender diversity, such as creativity and innovation". Regarding the size of the firm, some studies suggest that enterprises managed by women tend to be smaller than those run by men [114,115].
Logistic regression provides knowledge of the relationships among the variables, in this case, the propensity for firms to be run by women. Two logistic regression models were performed with Female CEO as the dependent variable.
A positive value of R 2 indicates that as the predictor variable increases, so does the likelihood of the CEO being women. A negative value implies as that the predictor variable increases, the probability of the outcome occurring decreases. If a variable has a small value of R 2 then it contributes only a small amount of the model. The odds ratio (Exp (B)) of an event occurring are defined as the probability of an event occurring divided by the likelihood of that event no occurring: If the value is higher than one, then it indicates that as the predictor increases, the odds of the outcome occurring increase. Conversely, a value less than one suggests that as the predictor increases, the odds of the outcome occurring decrease. Values less than one mean that as the predictor variable increases, the odds of CEO being women in 2013 decrease. The positive values indicate that the CEO will be a woman is more likely than will be a man, while the negative values indicate that be a woman executive is more likely than no to occur.

Results
In this section, the data analysis and SPSS results based on the Spanish SMEs data collected are analyzed and discussed. The empirical results are based on data for 522 service firms, in 2008; however, 97 data were lost in 2013, leaving the sample reduced to 425 companies. Regarding the follow-up loss bias, typical of cohort studies where observations are totally or partially lost over time, the values that can be considered subject to bias are not reached, since the value of 18.58% loss between the data of the first and the second cohort is below the value of 20% indicated in previous studies [116]. Table 2 summarizes descriptive statistics for the main variables analyzed in this study, the differences between the strategic decisions taken in 2008, differentiated by the female and male executives in 2013. Only 22.13 percent of executive managers are female over the sample period of 2013. In summary, the descriptive statistics show that more female executives run firms in 2013 that decided to grow by expanding the size of the business that held at that time by offering products or services related with the existing ones or expanding their geographical coverage. In addition, descriptive statistics indicate that firms that had introduced improvements in their products in the 2006-2008 period, had slightly above more female CEOs five years later, in 2013. The average age of the firms is 19.61 years in the case of men and 19.27 in the case of women; more female executives are found in younger companies than men executives. Regarding the size of the company measured by number of employees, the statistics' descriptive part shows that the average number of employees, in enterprises run by women was 3.54, while in firms run by men the average was 4.46 employees, so executive women manage more small firms than their male counterparts. Ownership of these companies was concentrated in few hands; in the case of companies run by male CEOs in 2013, the average was 3.58 owners, while businesses in companies directed by women had an average of 2.06 owners.
The main results of the two models of logistic regression in our investigation are presented in Table 3. From this information, research questions can be answered. Note: The levels of significance are * p < 0.1; ** p < 0.05; *** p < 0.01; **** p < 0.001; Categories of the gender of the CEO in 2013, 1: Female 0: Male 1 Two logistic regressions models were conducted to assess whether the independent variables significantly predicted whether or not a female executive run a firm. Being as the dependent variable, Female CEO is a dummy variable that takes the value one, when the CEO is a woman and zero otherwise. In the model one when all predictor variables are considered together, they significantly predict whether or not a woman runs a firm.
The full model one significantly predicted Female CEO (Omnibus chi-square = 31.57, df = 13, p < 0.03). The Chi-square p-value suggests that the overall model predicts female CEO occurrence or non-occurrence significantly better than when only the constant was included. The model accounted for between 24 percent and 40 percent of the variance on female CEO, with 84.6 percent of CEOs successfully predicted. Table 3 gives coefficients, the Wald statistic and associated degrees of freedom, and probability values for each of the predictor variables. This shows that items D14(1), D4(1), D5, and D6 reliably predicted female CEOs. The value of the coefficient of the position of the firm compared to the competition on the growing in recent years (item D6) reveals that an increase in the value of this predictor is associated with a decrease in the odds of female CEO, by a factor of 0.44. The coefficient of this variable suggests a negative relationship between this variable and female CEO occurrence. The implications of the results are as follows. They suggest that the position of the firm compared to the competition on the growing is negatively associated with female CEO, otherwise, it is more likely finding female CEOs in companies that have grown less than the competition in recent years. On the contrary, each unit increases the position of the firm compared to the competition introducing innovations (item D5) score is associated with an increase in the odds of female CEO by a factor of 2.41.
On the other hand, the categorical variable coding of D14(1), company plans to expand its activity in the coming years by expanding the geographical coverage, indicates that when the company plans to spread out its activity in the coming years expanding its geographic coverage, the probability that in the future the CEO being a woman is 4.37 times high than the probability of being a man. Furthermore, the item D4(1) suggests that when the firm applied any measure of reconciliation of work and family life, the likelihood that CEO being a woman in 2013 is 6.88 times greater than whether the company did not implement reconciliation measures.
The odds of female CEO improve by 2.41 times for each unit increase in D5 test score. Therefore, when the firm is more innovative than the competition, it is more likely that in the future the company is managed by a woman.
The Model II significantly predicted Female CEO (Omnibus chi-square = 40.12, df = 19, p < 0.03). The pseudo R 2 estimates indicate that approximately 30 percent and 51 percent of the variance on whether the CEO is a female can be predicted from the combination of independent variables. We can see that 88.6 percent of the CEOs were predicted correctly. Model II reports that D4 (1) The item D13(1) concerning the decision of the company to expand its activity in the coming years increasing the size of the current business is negatively associated with female CEOs, that is, it is more likely to find female CEOs in businesses that do not plan to expand their activity that in those that opt for growing. In contrast, the sign of D14(1) suggests that firms seeking to grow to expand the geographical coverage are 4.30 times more likely to be managed by women in the future.
The sign of the predictor D4 (1) indicates that firms that have adopted measures to reconcile work and family life are 15.76 times more likely to be managed by women in the future than those led by men.
The results suggest that when the relative position of the firm is less competitive in prices, it is more likely to have female executives, but the significance is low. On the other hand, when the position of the firm compared to its competitors on introducing innovations is stronger, the likelihood that the company will be administered by women is 9.7 times higher, with statistically significant differences at the level of p < 0.001. Notwithstanding, it is more likely to find female CEO in firms that have grown less than their competitors.

Discussion and Conclusions
This paper extends the work on female executives in the service sector. The study drew on a sample of SMEs in Spain to advance our understanding about the characteristics of businesses run by women in this industry. By looking at the 2008 global economic recession, the article attempts to investigate the contribution of strategic decisions of firms at that time in the future presence of female executives in SMEs. To this end, we formulated a questionnaire in 2008 about strategic business decisions and tried to predict the influence of these actions on the later likelihood that women ran companies.
Two binomial logistic regressions were conducted in order to find the set of predictors which best discriminate between male and female CEOs. The dependent variable is, therefore, a categorical variable which takes on the values zero or one depending on whether the CEO in 2013 has been female or not. In the first model, three statistically significant variables were selected in the logistic regression to explain female CEOs in SMEs in a positive way: firms that have affirmed they will expand their business in the coming years by expanding the geographical coverage, business that applied measures of reconciliation of work and family life, and firms that introduced more innovations compared to the competition. Chen et al. [88] also found that female representation is positively associated with performance only for leading innovative firms. Kanter [65] also noted that companies high on innovation tend to employ more women.
By contrast, a better position of the firm on growth terms, compared to competitors, has negatively contributed to predicting the existence of female executives. It is more likely finding female executives in companies that have grown less than their competitors in recent years. These findings are consistent with theories that suggest that women are less likely to take risks. This approach presumes that the theory of agency anticipates lower costs of agent behaviors contrary to the interests of the principal [31]. Regarding business that applied measures of reconciliation of work and family life, the results follow the same line as those obtained by Blasco-Burriel et al. [13] for Spanish women entrepreneurs. A possible explanation could be that the shortage of women entrepreneurs is due to the conflict between family and work life. Perhaps, the introduction of measures of reconciliation of work and family life in SME led women to accept an executive position. Flexibility between job and family would be in line with other studies [91,92]. This study also links up with the signaling theory that observes that the characteristics of the company, in this case of labor flexibility, could attract women to occupy executive positions. By incorporating women to the management position, companies contribute to reducing the glass ceiling and add to meeting the expectations required by society as demanded by the theory of legitimacy.
In the second model, the three statistically significant variables contributing to the prediction of businesses managed by women coincide with model I: firms that have affirmed they will expand their business in the coming years by expanding the geographical coverage, business that applied measures of reconciliation of work and family life, and firms that introduced innovations compared to the competition. In this model, two additional variables contribute significantly to predicting firms managed by women, the relative position of the firm on setting competitive prices influences positively, and the future growth of the company focusing on current products or markets impacts negatively. Again, the most risk-averse profile of women would explain this result.
Based on these results, it can be concluded that the probability that a CEO is a woman is higher on firms that at the height of the financial crisis affirmed that they applied measures of reconciliation of work and family life, were more innovative than their competitors, and would expand their business in the coming years by expanding the geographical coverage, but they have grown less than the competition in recent years and would not expand the size of their current business on the same products and markets. Han et al. [1] find that enterprise size has a significantly negative impact on the presence of female-CEOs.
The study proposes theoretical and practical insights into the corporate diversity debate. On the one hand, this article offers a further essential contribution by understanding female CEOs presence, predicting whether the CEO is male or female, and promoting sustainable small and medium enterprise development during times of crisis. Particularly, the results of the two models coincide in providing strong evidence that firms with a larger fraction of female CEOs (1) previously applied more measures of conciliation of labor and family life, (2) adopted growth strategy based on expanding the geographical coverage, and (3) were more innovative than their competitors. Our results also highlight that SMEs that have grown more than their competitors in recent years and have formulated strategies of growth intensifying the size of their current business are less likely to have female CEOs. These results are in line with those obtained by Han et al. [1], reflecting the influence of the firm characteristic on the gender of CEOs.
Moreover, our study contributes to complementary theoretical perspectives on corporate governance of SMEs firms. Several theories, such as agency, stakeholders, legitimacy, signaling or institutional approaches have been integrated in order to derive predictions about the relation between a number of strategic decisions of SMEs and the future presence of females on executive positions in the Spanish crisis context, as well as to obtain fuller understanding of and deep insights into female participation in business. We consider that it is not adequate to use a single theory for our theoretical framework due to the fact "that theory is always incomplete in the social sciences" [111] (p. 9).
The present study offers several significant and decisive contributions. With these results, this paper adds evidence to research showing that strategic decisions have both direct and mediating effects on the prediction of businesses run by women. Our analysis reveals that the future presence of female CEO is not so much influenced by past business decisions as by present strategies and by the forecasts about future policies to be followed in the medium and long term. As for current strategy, the commitment to innovation and conciliation as a competitive advantage over competing companies stands out for its positive effect on the future presence of women. In the opposite direction, it should be noted the strong position in terms of growth, since these firms will have the lowest future female presence. As for future strategies, among those companies that are committed to growth, the least risky strategy is these that have a more significant relationship with the female presence. The positive impact of growth through market diversification stands out, while future growth strategies based on the same products or services as the current ones, have the opposite effect.
According to our results, it can be suggested that companies and governments must continue to invest and promote innovation, as well as the implementation of conciliation measures, even in situations of difficulties and economic crisis, among other reasons because of their positive influence on the future presence of women on making-decision positions.
Therefore, our results are important, not only for academics and practitioners but for government and policy-makers. They should consider the importance of these findings in order to enhance the glass ceiling and gender diversity policies. They offer suggestions to improve the current regulation and practices. Guidelines focused on the mentioned issues will help to increase female participation in executive positions and to reduce the glass ceiling. In 2007, Spain became one of the first countries in the European Union to implement a recommended gender quota law for women's participation on boards [117]. Nowadays, large Spanish companies have to implement compulsory equality plans, and it is planned to extend to smaller companies.
Despite several significant findings, this study has some limitations, which suggest several opportunities for potential research. One of the limitations arises from the lack of more rich information about the profile of the CEO. Unfortunately, in the database used there is no information on some variables, such as the remuneration of women managers, age or previous work experience, and therefore, we could not perform some analysis. It should be taken into account that women are not randomly assigned o executive positions among firms, but unobservable individuals' characteristics may affect female promotion. In future research, it would be interesting to tackle this bias controlling for a set of individual executive features [118]. Nevertheless, the corporate governance perspective adopted in this analysis complements previous alternative approaches based on the CEO's profile.
On the other hand, we have combined primary information from a survey with data obtained from SABI (Iberian Balance Analysis System) database. The delay in the publication of the information led to the fact that more up-to-date analyzes could not be performed.
In the future, researchers could replicate the study in other regions or countries as well as in other industries and could even compare the service sector with other sectors with a lower presence of women in management positions. Differences in the economic, social or cultural context, as well as in the participation of women in the economic sphere, may imply differences in the results. It would also be interesting to analyze whether there are measures aimed at achieving greater flexibility between work and family in other sectors and countries. In addition, based on the agency's theory, in future studies, it would be interesting to analyze whether there is a duality between executive women and family members in SMEs so that by enjoying certain control rights over the company, they could influence the decision strategies in the organization.