In this study, we question: how do the social costs of failure interact with gendered institutions to affect the early stage entrepreneurship activity? We address this question by employing the institutional theory and a unique dataset of 286,989 entrepreneurs across 35 countries.
To test our hypotheses, we use a multilevel modeling analysis that nests individual entrepreneurs within the countries. To capture individual and country-level variables, we constructed a unique dataset that combines data from the Global Entrepreneurship Monitor (GEM), European Flash Barometer (EUFB), World Bank Development Indicator (WDI), World Bank Doing Business Report (WBDB) and World Economic Forum (WEF).
Our analysis confirms that higher levels of the country-level gender equality positively correlate with the early-stage entrepreneurship activity of women. Moreover, we find that this positive relationship is amplified in institutional environments with high social costs of failure, suggesting that societal intolerance for failure can exacerbate the negative effect of gender inequality on the participation of women in entrepreneurship.
Our research contributes to academic interest on the role of legitimacy in women entrepreneurship and is of particular interest to international business scholars, seeking a better understanding of multidimensional construction of institutional frameworks across countries. In this study, we set out to address an important research question: how do the social costs of failure interact with gendered institutions to affect entrepreneurship activity? Our study provides a comprehensive portrait of gendered institutions by including the framework conditions of education, healthcare and political power. We found that in societies with gender equality, the likelihood of individuals engaging in the early-stage entrepreneurship activity is higher and that the positive relationship is strengthened in national environments with high social costs of failure.
Our study findings underscore the need for government policies addressing global gender gaps in economic empowerment. In particular, policies assisting women in obtaining education in high-growth industries like information technology or providing funding to women-dominated industries may foster activity for women seeking to do business in such industries. Such policies connect the early-stage entrepreneurship activities with gender equality concerns and initiatives.
Regarding the social costs of failure construct, specifically, prior studies generally focus narrowly on the context of failed entrepreneurs. We cast a wider net on men and women entrepreneurs’ entry decisions (irrespective of prior experience with business failure) and provide new views on the effects of social costs of failure on entrepreneurial ecosystems. We also extend the research on the legitimacy of women as entrepreneurs with the gender equality construct.
Unlike previous studies, which often focus on the “3Ms” of market, money and management, our research adopts a more holistic perspective. We recognize that the opportunities and challenges faced by entrepreneurs are shaped not only by individual skills and resources but also by the broader macroenvironment. By incorporating the framework conditions of education, healthcare and political power, alongside the intricate interplay of social costs and norms, our study paints a comprehensive picture of the landscape of female entrepreneurship.
Introduction
Entrepreneurship thrives in supportive environments (Dheer et al., 2019). However, for women, social and institutional barriers often add to the liabilities of doing business that weaken venturing aspirations (Brush et al., 2017). This study delves into how gendered institutions, encompassing economic, political, educational and health dimensions, interact with societal perceptions of failure to influence the early-stage female entrepreneurship. Specifically, we explore how the high social costs of failure – societal punishments and judgments for nonconformity to stakeholder expectations – amplify the negative effect of inequality in gendered institutions on women’s entrepreneurial activity (Lee et al., 2021).
Gendered institutions are the ingrained norms and practices within economic, political, educational and health systems that shape opportunities and challenges for women. In many countries, the existence of gendered institutions is very evident from the gaps in labor force participation rates for women and in the wage inequality for similar work performed by women and men (World Economic Forum, Global Gender Gap Report, 2020). We argue that gendered institutional environments (Brush et al., 2017) affect the legitimacy of entrepreneurial efforts and, very importantly, the participation of women entrepreneurs in entrepreneurship activity (Wang et al., 2019).
We leverage the institutional theory (Scott, 1987, 1995; Suchman, 1995) and a rich dataset of 286,989 entrepreneurs across 35 countries to examine the effects of gender equality and the moderation of social costs of failure on women entrepreneurs. This approach allows us to capture the complex interplay between the societal expectations and institutional contexts using multi-item measures (Calza et al., 2020; Phillips and Zuckerman, 2001; Scott, 1987). Our analysis confirms that higher levels of the country-level gender equality positively correlate with the early-stage entrepreneurship activity of women. Moreover, we find that this positive relationship is amplified in institutional environments with high social costs of failure, suggesting that societal intolerance for failure can exacerbate the negative effect of gender inequality on the participation of women in entrepreneurship.
Our study contributes to several scholarly conversations on the effects of legitimacy and socialization contexts on entrepreneurship activity. First, by integrating the multi-item constructs of the social costs of failure and gender equality, we answer the call to move beyond single measures of the institutional drivers of the early-stage entrepreneurial activity (Lee et al., 2021). Second, we extend studies that call for more research on the legitimacy of women as entrepreneurs (Dheer et al., 2019; Brush et al., 2009). Third, our approach of moderating gender equality with social costs of failure informs the quality versus quantity debate on whether societies should motivate productive entrepreneurship activity using tactics that punish business failure (Lee et al., 2021; Simmons, 2012).
Unlike previous studies, which often focus on the “3Ms” of market, money and management (Brush et al., 2019), our research adopts a more holistic perspective. We recognize that the opportunities and challenges faced by entrepreneurs are shaped not only by individual skills and resources but also by the broader macroenvironment. By incorporating the framework conditions of education, healthcare and political power (World Economic Forum, Global Gender Gap Report, 2020), alongside the intricate interplay of social costs and norms, our study paints a comprehensive picture of the landscape of female entrepreneurship.
Literature review
The relationship between the institutional environment and the rate and types of entrepreneurship activity that emerges or succeeds has been a continuing subject of scholarly investigation. Of particular interest have been the effects of institutions on the entry and growth decisions of women entrepreneurs (Henry et al., 2016; Reutzel et al., 2023; Simmons et al., 2019). In many societies, the under-representation of women in economic, political, educational and health institutions has been shown to negatively impact the legitimacy and start-up activities of women entrepreneurs (Brush et al., 2017). In fact, negative social judgments and gender disparities are important concerns for countries focused on stimulating wealth and job creation through entrepreneurship activity (Matricano, 2022; Terjesen et al., 2016; Vracheva and Stoyneva, 2020).
Theories used in earlier studies to explain the participation of women entrepreneurs vary but generally center around institutional economics (see, e.g. Autio and Acs, 2010; Estrin et al., 2013a), cultural psychology (see, e.g. Autio et al., 2013) and gender role congruity (Reutzel et al., 2023; Vracheva and Stoyneva, 2020). Collectively, these prior studies find that systematic variation across cultures and social structures impact entrepreneurship (Cheraghi et al., 2019; Klyver et al., 2008; Simmons et al., 2019; Thomas and Mueller, 2000). Several of these studies have independently examined key factors such as economic and political influences (e.g. Vracheva and Stoyneva, 2020), specific contexts, such as high technology (e.g. Matricano, 2022; Reutzel et al., 2023) or the industry sector (e.g. Khan et al., 2021; Ringblom and Johansson, 2020). Other studies have attempted to include a broader scope of the institutional environment, focusing on women, immigrant and minority entrepreneurs (Fairlie, 2013), although many of these studies focus on market, money and management (“3Ms”) (Brush et al., 2019) and do not analyze the education, health and political factors of the macroenvironment.
Furthermore, the institutional theory perspective typically does not take into consideration that the firms within a specific institutional context are heterogeneous (Barney, 1991; Porter, 1980; Wernerfelt, 1984), such as men vs women-led entrepreneurial ventures, viewing the requirement for legitimacy as uniform. Indeed, an implicit assumption is that individual entrepreneurs embedded within a specific institutional environment will have comparable opportunities to start and grow ventures (Brush et al., 2019). However, the research has shown that heterogeneity in firm attributes can lead to heterogeneous legitimacy concerns (Young and Makhija, 2014; Reutzel et al., 2023), which, in turn, can cause variation in the impact of institutional pressures on entrepreneur’s decision-making. In this regard, gender equality in the institutional context plays an important role (Henry et al., 2016; Marlow, 2002) in the early-stage entrepreneurship activity.
Hypotheses development
According to Brush et al. (2019), gender can influence entrepreneurial ecosystems at the institutional level by setting boundaries for behaviors and actions (Welter and Smallbone, 2011; Smallbone and Welter, 2012). For instance, regulations that prohibit women from owning property (Hampel-Milagrosa, 2010) or moving freely about some countries (Estrin and Mickiewicz, 2011) can make collateralizing and establishing a business especially difficult. The same is true of informal institutions, often expressed as gender stereotypes and biases (Eddleston et al., 2016; Marlow and Swail, 2014; Vracheva and Stoyneva, 2020). In addition, gendered social roles allocate different responsibilities to men and women that influence their career perceptions (e.g. Lent et al., 2002; Reutzel et al., 2023). Conformity to social roles impacts the level of stakeholders’ trust and hence, the likelihood of starting a business (Dheer et al., 2019; Reutzel et al., 2023).
As such, gender equality will have an influence on entrepreneurial behaviors (Eddleston and Powell, 2008; Gupta et al., 2009). Engagement in entrepreneurship has been shown to increase in the institutional settings with gender-blind regulatory frameworks and government policies (Wang et al., 2019) and where entrepreneurs perceive themselves to fit into the context (Hsu et al., 2019; Yacus et al., 2019). This is important because in environments, where social roles exclude women from participation in economic participation and opportunity, educational attainment, health and survival and political empowerment, they are shown to form different career trajectories (e.g. Lent et al., 2002; Ringblom and Johansson, 2020). Women and men will also form different perceptions of their ability to navigate selection pressures (Fisher et al., 2016; Miner et al., 1990) and attract the human, financial and social capital (Garud et al., 2014) needed to start and survive in new firms (Canning et al., 2012; Greene et al., 2003). The outcome is likely to be that fewer women will enter entrepreneurship. We, thus, argue that gender equality increases the overall supply and engagement of entrepreneurs. We suggest:
Gender equality at the national level increases the likelihood of early-stage entrepreneurship activity.
Moderating effects of social costs of failure
Examples of high social costs of failure include the judgments or actions of stakeholders that lead to constrained access to human, social and financial resources (EOS Gallup Europe, 2010). Social costs of failure are often formally framed into laws, regulations and procedures that have unfriendly consequences for stigmatized entrepreneurs (Kanze et al., 2020; McMullen et al., 2008). High social costs of failure signify societal intolerance for entrepreneurs with spoiled social identities (Goffman, 1963) and for entrepreneurial behaviors that deviate from expected social norms (Lee et al., 2007; Simmons et al., 2014; Sutton and Callahan, 1987; Ucbasaran et al., 2013). Independently, high social costs of failure have been shown to have a negative effect on the early-stage entrepreneurial activity (Lee et al., 2021).
As a moderating variable, high social costs of failure signal a stringent selection process for inclusion into the entrepreneurial ecosystem. Accordingly, we argue that there will be gender differences in the effects of social costs of failure on entrepreneurial behavior because women and men have different concerns with respect to social judgments and punishments for nonconforming behaviors and outcomes (Simmons et al., 2019; Van der Zwan et al., 2016).
In national environments with low levels of gender equality, women face greater risks of nonconformity, since prescribed social roles exclude them from equal participation in economic opportunity, educational attainment, health and survival and political empowerment (Brush et al., 2017; Simmons et al., 2014). As such, women face greater social costs of failure, which include the anticipated consequences for nonconformity than men (Lee et al., 2021; Rafiullah et al., 2023). In these gendered environments, entrepreneurship is more likely to be viewed as a masculine social role that is unsuitable for women (Brush et al., 2017; Simmons et al., 2019).
This greater social cost of failure represents an additional risk of prospective women entrepreneurs. Since women are generally more risk averse than men (Croson and Gneezy, 2009), this additional risk will deter greater numbers of women from engaging in the early-stage entrepreneurial activity than men. However, in nations with greater gender equality, the social roles and expectations of women and men are more aligned (Brush et al., 2017; Simmons et al., 2014). Thus, the social cost of failure in entrepreneurship activity for women should be more like that for men (Rafiullah et al., 2023; Simmons et al., 2019). Since the social cost of failure for engaging in entrepreneurial activity is more similar for both genders, we expect that the rate at which women pursue entrepreneurship to be more like that of men.
In short, the negative effects of the social cost of failure on the early-stage entrepreneurship activity of women entrepreneurs are strengthened by gender inequality and weakened by gender equality in national environments (see Figure 1). Therefore, as the levels of social costs increase, the disparity in the entrepreneurship activity between men and women is also likely to increase in national environments, where women also must consider the negative social judgment of their participation in economic, political, educational and health institutions. This argument aligns with prior research, which shows that entrepreneurs with less favorable perceptions of their institutional environment are less likely to be opportunity focused entrepreneurs (Van der Zwan et al., 2016). An experiment by Gupta et al. (2019) demonstrates the relevance of this need for fit, finding that women consider the social context in their cognitive choice to pursue entrepreneurship activity. We suggest:
As the social cost of failure increases, the positive relationship between gender equality and the probability of engagement in the early-stage entrepreneurship activity at the individual level is strengthened.
Methodology
Research design and sample
To test our hypotheses, we use a multilevel modeling analysis that nests individual entrepreneurs within the countries. To capture the individual- and country-level variables, we constructed a unique dataset that combines data from the Global Entrepreneurship Monitor (GEM), European Flash Barometer (EUFB), World Bank Development Indicator (WDI), World Bank Doing Business Report (WBDB) and World Economic Forum (WEF).
The individual-level data were taken from the GEM Adult Population Survey (APS). The GEM project conducts an ongoing cross-national survey with the purpose of measuring entrepreneurial activities across countries (Bosma, 2013). To increase the stability of the measures, we pooled the GEM data across the four-year period of 2009–2012 and included only the working population between the ages of 18 and 64.
Country-level variables were drawn from the EUFB, WDI, WBDB and WEF. The EUFB #283 and #354 reported public attitudes toward entrepreneurship, such as entrepreneurial education, risk-taking, obstacles to entrepreneurship and business failure for European countries and beyond (EOS Gallop Europe, 2010, 2013). The WDI database is a public dataset published by the World Bank organization, which shows national development indicators including GDP per capita, GDP growth, and population size (World Bank Group, 2018). The World Economic Forum (WEF) provides various data regarding the issues of the world economy to business leaders, international political leaders and intellectuals, including the gender gap in each country (World Economic Forum, 2012). Our final dataset consists of 286,989 GEM respondents between the ages of 18 and 64 from 35 countries. We summarize the variables and present descriptive statistics and correlations in Tables 1 and 2, respectively, and present the list of countries in Table 3.
Definitions of variables
| Level | Variable | Definition | Source |
|---|---|---|---|
| Country | Gender equality | The level of gender equality | WEF Gender Gap Report (2009–2012) |
| Social costs of failure | Stigma of bankruptcy + The depth of credit information (Coverage, scope and accessibility of credit information) | Flash Euro Barometer #257 and #354+ World Bank Doing Business (2008–2011) | |
| per capita GDP ppp (t‒1) | per capita GDP at purchasing power parity at 2005 $USD (natural log) | World Bank WDI (2008–2011) | |
| GDP growth rate (t‒1) | GDP growth rate per year | World Bank WDI (2008–2011) | |
| Population size (t‒1) | Total number of population (log) | World Bank WDI (2008–2011) | |
| Individual | Early-stage entrepreneurship | 1: individuals engaged in nascent entrepreneurial activity or operating a venture less than 42 months; 0 otherwise | GEM APS (2009–2012) |
| Age | Age of respondents (Min = 18, Max = 64) | GEM APS (2009–2012) | |
| Income | Household income level one (lowest 33%), two (middle 33%) and three (top 33%) | GEM APS (2009–2012) | |
| Education | Respondent’s education: four (graduate experience), three (post-secondary) and two (secondary). one (no secondary degree) | GEM APS (2009–2012) | |
| Gender | Respondent’s gender: 1 (male) and 0 (female) | GEM APS (2009–2012) |
Source(s): Table by authors
Descriptive and correlations for individual- and country-level variables
| Mean | Std. Dev. | (1) | (2) | (3) | (4) | (5) | ||
|---|---|---|---|---|---|---|---|---|
| Individual-level correlations | ||||||||
| (1) | Early-Stage Entrepreneurship | 0.07 | 0.25 | 1 | ||||
| (2) | Age | 41.91 | 12.93 | −0.06 | 1 | |||
| (3) | Income | 2.20 | 0.80 | 0.05 | −0.03 | 1 | ||
| (4) | Education | 2.20 | 1.01 | 0.04 | −0.10 | 0.28 | 1 | |
| (5) | Gender | 0.48 | 0.49 | 0.07 | −0.03 | 0.09 | 0.01 | 1 |
| Mean | Std. Dev. | (1) | (2) | (3) | (4) | (5) | (6) | ||
|---|---|---|---|---|---|---|---|---|---|
| Country-level correlations | |||||||||
| (1) | Early-stage entrepreneurship | 0.07 | 0.25 | 1 | |||||
| (2) | Log GDP PPP | 9.99 | 0.80 | −0.08 | 1 | ||||
| (3) | GDP growth rate | 0.52 | 3.90 | 0.08 | −0.41 | 1 | |||
| (4) | Population size | 17.32 | 1.57 | 0.04 | −0.36 | 0.35 | 1 | ||
| (5) | Social costs of failure | −0.15 | 1.11 | −0.06 | 0.45 | −0.14 | −0.2 | 1 | |
| (6) | Gender equality | 0.72 | 0.04 | −0.03 | 0.59 | −0.24 | −0.36 | 0.04 | 1 |
Note(s): All significant at 0.05 level
Source(s): Table by authors
Country sample description
| Country | OECD membership | Gender equality | Social cost of failure | % early-stage entrepreneurship (TEA) out of total population |
|---|---|---|---|---|
| Austria | Yes | 0.74 | 0.80 | 9.58 |
| Belgium | Yes | 0.75 | 0.09 | 4.52 |
| Brazil | No | 0.67 | −1.72 | 15.79 |
| China | No | 0.69 | −1.21 | 17.51 |
| Croatia | No | 0.70 | 0.35 | 6.14 |
| Czech Republic | Yes | 0.68 | 0.91 | 7.64 |
| Denmark | Yes | 0.77 | 0.37 | 4.35 |
| Estonia | Yes | 0.70 | −0.22 | 14.26 |
| Finland | Yes | 0.83 | −0.15 | 5.78 |
| France | Yes | 0.71 | −0.45 | 5.27 |
| Germany | Yes | 0.75 | 1.03 | 4.81 |
| Greece | Yes | 0.68 | −0.70 | 7.19 |
| Hungary | Yes | 0.67 | 1.41 | 7.94 |
| Iceland | Yes | 0.84 | 1.36 | 11.02 |
| India | No | 0.64 | 1.16 | 10.69 |
| Ireland | Yes | 0.78 | −0.61 | 6.72 |
| Israel | Yes | 0.70 | 0.70 | 5.87 |
| Italy | Yes | 0.68 | 0.56 | 3.46 |
| Japan | Yes | 0.65 | 1.35 | 3.94 |
| South Korea | Yes | 0.63 | 1.10 | 7.01 |
| Latvia | Yes | 0.75 | −1.24 | 10.68 |
| Lithuania | Yes | 0.71 | 0.11 | 8.98 |
| The Netherland | Yes | 0.75 | 1.10 | 8.23 |
| Norway | Yes | 0.84 | 0.33 | 7.49 |
| Poland | Yes | 0.70 | 0.41 | 9.20 |
| Portugal | Yes | 0.71 | 0.16 | 6.54 |
| Romania | No | 0.68 | −0.47 | 6.40 |
| Russia | No | 0.70 | −0.39 | 4.18 |
| Slovakia | Yes | 0.68 | −0.10 | 12.21 |
| Slovenia | Yes | 0.71 | −0.60 | 4.55 |
| Spain | Yes | 0.74 | 0.13 | 5.23 |
| Sweden | Yes | 0.81 | −0.90 | 5.71 |
| Switzerland | Yes | 0.76 | 0.74 | 6.32 |
| UK | Yes | 0.74 | 0.62 | 7.11 |
| USA | Yes | 0.73 | 0.46 | 10.18 |
Source(s): Table by authors
Variables
Dependent variable
Early-stage entrepreneurial activity is constructed from the Global Entrepreneurship Monitor Adult Population Survey (GEM APS) dataset (Autio et al., 2013; Stephan and Uhlaner, 2010). Early-stage entrepreneurial activity is a combination of nascent entrepreneurs currently involved in starting a new business and the owners of young businesses in operation less than 42 months, and this variable is measured dichotomously, with 1 indicating that the individual is engaged in early-stage entrepreneurial activity.
Independent variable
Gender equality is an institutional variable at the country-level, constructed by utilizing the Global Gender Gap Index (GGGI) from the WEF Gender Gap report for 2009–2012. Prior studies have utilized GGGI from the WEF report in the entrepreneurship context to measure the country-level gender equality differences (Brush et al., 2017; Nagy, 2006; Rouse et al., 2013). In our study, the higher gender equality score is indicative of a higher level of the representation of women in economic, political, educational and health institutions.
The GGGI variable from the WEF report has four dimensions: economic participation and opportunity, educational attainment, health and survival and political empowerment. Economic participation and opportunity measures labor force participation, wage equality for similar work and estimated earned income (World Economic Forum, 2020). Educational attainment measures literacy and enrollment in primary education, secondary education and tertiary education. Health and survival measures sex ratio at birth and healthy life expectancy data from the United Nations, Department of Economic and Social Affairs and World Health Organization (World Economic Forum, 2020). Political empowerment measures women in parliament, ministerial positions, and head of the state data from the Inter-Parliamentary Union (IPU), (2019) and (World Economic Forum, 2020).
Moderating variable
Social costs of failure is an institutional-level variable that refers to the negative consequences faced by the entrepreneurs, who have experienced business failure and are impacted by the societal stigma associated with failure and the levels of visibility of their business failure (Simmons et al., 2014). As validated by Lee et al. (2022), this variable is constructed by combining the standardized score of the stigma of failure and the standardized score of depth of credit information. Specifically, this variable integrates the informal institutional norm of the stigma of business failure and the formal institutional regulation of getting credit (Lee et al., 2022; Simmons et al., 2014). We use the EUFB #283 (EOS Gallop Europe, 2010) from a survey circulated in 2009 and #354 (EOS Gallop Europe, 2013) from a survey collected in 2012, which measured the country-level attitudes in the European Union between 2009 and 2012. The EUFB questions measure societal attitudes about doing business with entrepreneurs who experience business failure. The formal institutional variable constructed for the depth of credit information captures the visibility of entrepreneurial failures (World Bank Group, 2013, 2018; see Lee et al., 2021). “Depth of credit information index measures rules affecting the scope, accessibility and quality of credit information available through public or private credit registries. The index ranges from 0 to 8, with higher values indicating the availability of more credit information, from either a public registry or a private bureau, to facilitate lending decisions” (World Bank, Doing Business project, doingbusiness.org).
Control variables
This study controls for variables on two levels: individual and institutional. As individual differences can influence entrepreneurial decisions (Shepherd et al., 2015), we control for age, education, gender and household income at the individual-level, taken from the GEM APS dataset. Age of the entrepreneur is measured as a continuous variable between 18 and 64 years, since age may influence the entrepreneurial activity (Liang et al., 2018). Research has also shown that human capital has a positive relationship to both new venture growths (Baum and Bird, 2010), and the effectiveness of institutional policies aimed at reducing barriers to high-growth entrepreneurship (Eesley, 2016). We, thus, control for the entrepreneur’s education, measured in four categories: (1) no secondary degree, (2) secondary degree, (3) post-secondary degree and (4) greater than graduate-school degree. As financial capital determines entrepreneurial decisions, we control for household income, a categorical variable assessing whether a respondent belongs to the (1) lower, (2) middle or (3) higher tier of a country’s distribution of household income. Finally, we control for gender, a binary variable: 1 represents male, whereas 0 represents the female.
This study also controls for the effects of several institutional-level factors. The degree of a country’s development influences the rate of entrepreneurial entry (Acs et al., 2008). Hence, we control the economic development of a country by using a natural log of per capita GDP at purchasing power parity (GDP PPP). Second, as a country’s GDP growth rate is correlated with entrepreneurship activity (Acs et al., 2008; Estrin et al., 2013a, b), we include GDP growth rate. Finally, we control for population size, as the size of the market can be influenced by the population size of each country, which can play an important role as an institution for entrepreneurial activities (Sato et al., 2012).
Statistical approach
We conducted a series of multilevel random effects logistic regressions. This method is appropriate, given that individuals within the countries share common experiences that differ from those living in other countries (Davidsson and Wiklund, 2001; Estrin et al., 2013a, b; Stephan et al., 2015) and is consistent with recent uses of multilevel analysis in entrepreneurship and international business studies (Estrin et al., 2013b; Lee et al., 2022; Stephan et al., 2015). We first conducted an intra-class correlation (ICC) analysis to justify using a multilevel regression model. The multilevel techniques are recommended if the ICC estimates reside within the normal range (i.e. between 5 and 20%) (Bliese, 2000). Our results indicate that 5.6% of the total variance for general entrepreneurial activities resided at the country level. In our multilevel regression model, we also examine the variance inflation factor (VIF) statistics to control the possibility of strong multicollinearity influencing our results. We find that all VIF scores are below 10, suggesting that multicollinearity is not a concern for our analysis (Hair et al., 1998). To aid the interpretation, we present our results with odds ratios (OR) in lieu of log-odds coefficients.
Results
Before testing our hypotheses, we ran regressions with only control variables. Model 1 in Table 3 shows that a higher probability of individual-level engagement in the early-stage entrepreneurship activity is positively associated with an individual who is more likely to be younger, wealthier or have a higher level of education. Furthermore, we find that a country’s wealth level is negatively associated with the probability of individual-level engagement in the early-stage entrepreneurship activity. We did not hypothesize the direct effects of social costs of failure on the early-stage entrepreneurial entry because these relationships have already been tested and supported in the prior literature (Lee et al., 2021).
Because we tested the moderation effects of social costs of failure in Hypothesis 2, we examined the direct effects of the variable to validate the consistency with the prior research (Lee et al., 2021). We find a statistically significant negative association between the social cost of failure at the country-level and the probability of individual-level engagement in the early-stage entrepreneurship, shown in Model 2 in Table 4 (OR = 0.97, p < 0.1), aligned with Lee et al. (2021).
Results of logistic multi-level regression for early-stage entrepreneurship
| Model 1 | Model 2 | Model 3 | |
|---|---|---|---|
| O.R. | O.R. | O.R. | |
| Fixed part | |||
| Individual level (control) | |||
| Age | 0.983*** | 0.983*** | 0.983*** |
| (0.001) | (0.001) | (0.001) | |
| Income | 1.219*** | 1.219*** | 1.218*** |
| (0.012) | (0.012) | (0.012) | |
| Education | 1.169*** | 1.169*** | 1.169*** |
| (0.010) | (0.010) | (0.010) | |
| Gender | 1.729*** | 1.729*** | 1.729*** |
| (0.026) | (0.026) | (0.026) | |
| Country level (control) | |||
| GDP PPP | 0.671┼ | 0.662*** | 0.679*** |
| (0.069) | (0.064) | (0.059) | |
| GDP growth rate | 1.004 | 1.002 | 1.001 |
| (0.004) | (0.004) | (0.004) | |
| Population size | 0.899* | 0.918┼ | 0.949 |
| (0.044) | (0.043) | (0.041) | |
| Social costs of failure | 0.968* | 0.253*** | |
| (0.015) | (0.087) | ||
| Country level (independent) | |||
| Gender equality | 6.727* | 21.20*** | |
| (5.596) | (18.47) | ||
| Moderating effect | |||
| Gender equality*Social cost of failure | 6.267*** | ||
| (2.961) | |||
| Random part and model fit | |||
| Intercept | 13.989┼ | 2.878 | 0.558┼ |
| (21.955) | (4.491) | (0.068) | |
| Deviance | 142041.37 | 142023.24 | 142013.49 |
| Wald χ2 | 3662.34 | 3683.61 | 3695.13 |
| Prob > χ2 | 0.00 | 0.00 | 0.00 |
| LR Test Prob < χ2 | 0.00 | 0.00 | 0.00 |
| # of Observation | 286,988 | 286,988 | 286,988 |
| # of Countries | 35 countries | 35 countries | 35 countries |
| Obs. per group min. | 1,480 | 1,480 | 1,480 |
| Obs. per group avg. | 8,199 | 8,119 | 8,119 |
| Obs. per group max. | 58,128 | 58,128 | 58,128 |
Note(s): p-Value: ***p < 0.001, **p < 0.01, *p < 0.05 and ┼p < 0.1. SE Value: Indicated within the parenthesis
Source(s): Table by authors
Hypothesis 1 states that gender equality at the national-level increases the likelihood of the early-stage entrepreneurship activity at the individual-level. As shown in Model 2 in Table 4, we find a statistically significant positive relationship (OR = 6.73, p < 0.05) between the level of gender equality and the likelihood of an individual to engage in the early-stage entrepreneurship. The ICC change can be interpreted as an effect size for the multilevel analysis (Lorah, 2018). Gender disparity and the social cost of failure explained 16.1% of the country-level variance. This result supports Hypothesis 1.
Hypothesis 2 states that the positive relationship between gender equality and the probability of engagement in the early-stage entrepreneurship activity at the individual level is strengthened in institutional environments with high social costs of failure. The result in Model 3 in Table 4 shows that there is a significant positive moderating effect of the social cost of failure on the relationship between the levels of gender equality at the country-level and the probability of engagement in the early-stage entrepreneurship activity at the individual-level (OR = 6.27, p < 0.001), while increasing 1.2% of additional explanation at the country-level variance. This suggests that when a higher level of social costs of failure exists in a country, the positive relationship between the levels of gender equality and the probability of engagement in the early-stage entrepreneurship activity at the individual-level is strengthened. Thus, Hypothesis 2 is supported (see Figure 2).
Robustness tests
We conducted additional analyses. First, we included only OECD member countries in our sample. The results were very similar to our main analysis. We ran another regression, by excluding individuals who experienced business failure within a year. The result does not differ significantly from the findings of our main analysis. We also examined the independent effects of the four dimensions of the WEF gender equality. This ad hoc test showed a significant finding for the moderating effects of gender equality in economic participation and opportunity and social costs of failure. Finally, we conducted analyses with samples of men and women only. We found significant direct effects of gender equality on the probability of engagement in the early-stage entrepreneurship activity as well as significant moderating effects for social costs of failure.
Discussion
Our research contributes to academic interest on the role of legitimacy in women entrepreneurship and is of particular interest to international business scholars seeking a better understanding of multidimensional construction of institutional frameworks across countries. In this study, we set out to address an important research question: how do the social costs of failure interact with gendered institutions to affect entrepreneurship activity? Our study provides a comprehensive portrait of gendered institutions by including the framework conditions of education, healthcare and political power (World Economic Forum, Global Gender Gap Report, 2020). We found that in societies with gender equality, the likelihood of individuals engaging in the early-stage entrepreneurship activity is higher and that the positive relationship is strengthened in national environments with high social costs of failure.
In societies with lower gender equality, women face greater social judgment and punishment for engaging in entrepreneurship activities (Brush et al., 2019). If the social costs of failure are high, then women faced with gender inequality could perceive additional risks above that of their male counterparts. Therefore, these women are less likely to pursue the early-stage entrepreneurship activity, given the general higher level of risk aversion of women entrepreneurs (Croson and Gneezy, 2009).
In this study, we incorporate the multidimensional constructs of the social costs of failure and gender equality (Lee et al., 2021). As scholars, we have more to learn about the effects of cultural institutions on driving the engagement of women entrepreneurs into entrepreneurial projects that have high-growth potential (Shepherd and Patzelt, 2017). At present, the literature focuses more on one-dimensional deterrents, such as the lack of access to childcare in the social sphere (Terjesen et al., 2016) or the lack of access to external funding in the economic sphere (Thébaud, 2010; Bird and Brush, 2002; Gupta et al., 2009). Our study addresses the calls in the literature for a better understanding of the more complex institutional drivers that lead to new or heightened forms of gender inequalities (Ukhova, 2015) or that push entrepreneurs to and from entrepreneurial projects (Shepherd and Patzelt, 2017) in different countries.
Regarding the social costs of failure construct, specifically, prior studies generally focus narrowly on the context of failed entrepreneurs (Simmons et al., 2014). We cast a wider net on men and women entrepreneurs’ entry decisions (irrespective of prior experience with business failure) and provide new views on the effects of social costs of failure on the entrepreneurial ecosystems. We also extend the research on the legitimacy of women as entrepreneurs with the gender equality construct (Dheer et al., 2019). For instance, Brush et al. (2009) argues that in the institutional contexts, where women are unable to socialize or engage equally in the economy, they have limited decision choices, search processes and perceptions about the existence or accessibility of entrepreneurial opportunities. By exploring the relationship between the country-level gender equality and individual decisions to engage in the early-stage entrepreneurship activities, our study responds to Brush et al. (2009)’s call for research on the effects of national and regional level policies, culture, laws, economy and support services on the exercise of choice for women entrepreneurs.
Practitioner and policy implications
Research has shown that public policy decisions can have gendered effects on engagement in entrepreneurship generally (Terjesen et al., 2016). Despite decades of effort on policy reforms aimed at reducing global gender gaps in economic empowerment, especially regarding women accessing necessary entrepreneurship funding (Guzman and Kacperczyk, 2019; Kanze et al., 2020), few countries have closed at least 50% of their gender gaps for economic participation and opportunity (World Economic Forum, Global Gender Gap Report, 2020). Women are also persistently less present in labor markets, with only 18.2% of firms being woman-led globally and 72 countries, where some women cannot access credit or open bank accounts (World Economic Forum, Global Gender Gap Report, 2020).
Our study findings underscore the need for government policies addressing global gender gaps in economic empowerment (Guzman and Kacperczyk, 2019; Shepherd and Patzelt, 2017). Policies assisting women in obtaining education in high-growth industries like information technology or providing funding to women-dominated industries may foster activity for women seeking to do business in such industries (Guzman and Kacperczyk, 2019). Such policies connect the early-stage entrepreneurship activities with gender equality concerns and initiatives (Dheer et al., 2019).
Study limitations
Our study has limitations, which we invite future research to address. First, while we examine our hypotheses with a sample of more than 286,000 individuals from 35 countries and control several country-level institutions such as GDP per capita, GDP growth rate and population size, we limited our sample to developed countries for comparability. As there are differences between developed and less-developed countries for entrepreneurship activities (Hessels and van Stel, 2011); we join in the call for more studies on the entrepreneurship activities and institutions in emerging economies (Urbano et al., 2020).
Second, while our study findings help us to better understand the unique role that societal attitudes play in moderating entrepreneurial behavior (Meek et al., 2010), future research could take this further, by integrating the social costs of failure construct into a study of the effect of cultural dimensions on entrepreneurial activities. Scholars could explore how dimensions, such as individualism-collectivism or power distance (Hofstede, 1980, 2011; House et al., 2004) interact with social costs of failure in different firm endeavors. This exploration could go beyond entrepreneurship, delving into cultural effects on firms’ corporate social responsibility and sustainability efforts, knowledge transfer across subsidiaries of MNEs or entry mode.
Third, future research could also explore the impact of the social cost of failure on high growth entrepreneurship, an important topic due to the impact of this form of entrepreneurial activity on economic growth and job creation (Henrekson and Johansson, 2010; Shane, 2009). Prior studies demonstrate that the social attributes of institutions can foster or discourage high-growth entrepreneurship (Shapero and Sokol, 1982; Shneor et al., 2016). Therefore, the examination of the effect of institutions on entrepreneurial activity is a worthwhile endeavor, as prior work has shown that informal institutions such as controlling corruption, having confidence in one’s skills and the ability to access credit have an impact on opportunity entrepreneurship (i.e. high-growth aspirations), which in turn, promotes economic growth (Aparicio et al., 2016).
Conclusion
This study used an institutional framework to examine the effects of gender equality and social costs of failure on the early-stage entrepreneurial activity and found evidence that these framework conditions can affect the quantity and quality of entrepreneurship activity in a region. Just as a rising tide can lift all boats, addressing barriers to women participating in the entrepreneurial activity and having gender equality in the representation of women in economic, political, educational and health institutions can benefit economies worldwide.
