Social Entrepreneurial Ecosystems in Upper-Middle-Income Countries: Social Policy and Sustainable Economic Development Implications

Abstract

This study aims to explore how a region’s degree of vulnerability influences the perceived level of support toward social entrepreneurship from a social entrepreneurial ecosystem perspective. The study of social entrepreneurial ecosystems (SEEs) constitutes a promising area for its potential to support high-impact entrepreneurs, but they are currently underexplored in upper-middle-income countries. This study also combines a macro-level (countries) and a meso-level (vulnerability regions) analysis to better understand how national policies may affect the perceptions of different ‘regions’ and, consequently, sustainable economic development. This study follows principal component analysis and non-parametric statistics to compare the means of the countries and the levels of regions’ vulnerabilities based on the Global Entrepreneurship Monitor’s regional data from Chilean (n = 276) and Mexican (n = 188) experts. At the macro level, the results show that experts in Chile have a better perception of social policies for social entrepreneurship than their counterparts in Mexico. At the meso-level, experts in high-vulnerability regions have a better perception of the social ecosystem dynamism than those in the low-vulnerability regions. The results contribute to the urgent dialogue to set up strategies that allow upper-middle-income countries and their regions to achieve greater impact and sustainability.

1. Introduction

Billions of people continue to live in poverty and are denied a decent living [1]. Social entrepreneurs are drivers of change and respond to these social challenges by providing social innovations [2,3]. They combine the efficiency and innovation of profit-making enterprises with the passion, values, and purpose of third-sector organizations [4,5] to empower disadvantaged groups. They also aim to scale up social innovations and solutions beyond local contexts to generate more impact [6,7,8]. Therefore, social entrepreneurs need support networks to help them finance, design, implement, and scale their initiatives to generate structural changes, new equilibriums, and shifts in power dynamics [9]. Thus, social entrepreneurship is considered a source of appropriate innovations in social policy delivery and financing [3] and an effective strategy to achieve sustainable development goals (SDGs) [10]. The term “Sustainable Economic Development”, by itself, refers to economic development aimed at balancing economic growth with environmental and social sustainability. Therefore, for this study, we argue that social entrepreneurship is a tool of social policy that is widely used to reach social sustainability.

Bretos et al. [11] mentioned that a promising research avenue relies upon examining how regional ecosystems affect social enterprises’ scaling decisions, processes, and outcomes from an institutional approach. Sinclair et al. [12] found that the structural conditions and institutional factors beyond their control constrain the impact of social enterprises and social innovations. The entrepreneurial ecosystem approach presents a systemic perspective of how to support entrepreneurial activities [13]. This approach benefits key actors, such as social entrepreneurs, that deal with complex systems of interacting players in rapidly evolving environments [14]. Traditional entrepreneurial ecosystems (EEs) may have conditions that can promote the creation and development of social ventures, although their objective is not primarily social but economical [15,16]. For instance, Lepoutre et al. [17] found that countries with higher rates of traditional entrepreneurial activities also have higher rates of social entrepreneurial activities. However, social enterprises have different logic and goals [2] and establish more cross-sector partnerships [18] than enterprises seeking profit maximization. While public authorities are willing to finance interventions to address social risks, such as social enterprises, they are much less inclined to design policies that involve them directly in the delivery [3]. Consequently, social investment and social entrepreneurship are linked to the narrative of social policy. Policymakers constantly promise to improve the conditions of their EEs to foster social cohesion and support for entrepreneurs.

Hon Chan et al. [19] argued that creating an enabling environment for social entrepreneurship in tackling complex socioeconomic challenges is at the global forefront of government policy agendas. However, there is scarce literature on the social entrepreneurial ecosystem (SEE), resulting in an unclear role for social entrepreneurs [20]. Still, researchers and policymakers are concerned about developing supportive ecosystems that promote social innovation initiatives [21]. Researchers are beginning to formally explore how social and conventional entrepreneurial ecosystems are distinct phenomena [22]. The sustainable entrepreneurial ecosystem is the closest concept in the ecosystem’s literature exploring this [23,24]. Nevertheless, it is differentiated from the social entrepreneurial ecosystem concept because a sustainable entrepreneurial ecosystem implies integration with environmental, social, and governance objectives aligned with the United Nations (UN) sustainable development goals (SDGs) [23,25,26], while the social entrepreneurial ecosystem overlaps with addressing SDGs but primarily aims to create social impact through social entrepreneurship and does not address governance objectives.

In summary, social entrepreneurial ecosystems focus on the social impact outcome, and sustainable entrepreneurial ecosystems concentrate on the management process of reaching the desired long-term socioeconomic outcome. Both are relatively new and underexplored concepts in the literature. Another venue is the research suggesting that community engagement is an essential driver of social EEs [27]. The community is often a taken-for-granted social group category in the social scientific literature, characterized by ongoing face-to-face interaction, spatial proximity, and shared institutions such as religious institutions, kinship, or schools [28,29]. Moore [30] argued that the members of an ideal ecosystem are motivated to work together to benefit the community. However, instead of analyzing the broader concepts, we preferred to deeply understand the social component of sustainability using the theory of a social entrepreneurial ecosystem (SEE) using a well-defined measurement in a new context, while the others remain unclear on the proper way of studying them.

On this pathway, Leme et al. [31] found that many social business ideas in Latin America are undiscovered. These social entrepreneurs fail to evaluate their mission’s social impact [32]. Moreover, Latin America’s social investment or impact investing market is in the early stages of development [33]. Impact investing is a general investment strategy seeking to generate financial returns while creating a positive social or environmental impact. The actors are still figuring out how to promote structural changes that will help consolidate SEEs. Additionally, the evidence shows that a weak EE and the absence of entrepreneurship support policies make social entrepreneurs rely almost exclusively on their innovative capacity [34]. For ecosystems to support social entrepreneurship, practitioners, scholars, governments, and civil societies should consider different types of innovation, particularly those that go beyond technology [35]. Also, policymakers from other countries have turned to fostering and supporting actors such as social entrepreneurs that they hope will achieve social inclusion and cohesion [3].

Fehrer et al. [36] argued that SEEs comprise multiple actors and each of them is active in one or several smaller ecosystems that are organized around solving specific social issues like a community. Since these social issues are usually located in high-vulnerability (HV) regions, the availability of social policies and dynamic networks might be particularly valuable for social entrepreneurs in these regions [16]. Hence, the objective of this study is to explore how a region’s degree of vulnerability influences the perceived level of support toward social entrepreneurship from a social policy and social ecosystem dynamism perspective in upper-middle-income countries. This study also argues that it is essential to build and reinforce SEEs as a strategy to address the SDGs and contributes to describing the hybrid organization and structure of entrepreneurial ecosystems [20,26,37].

This paper is structured as follows. The next section provides a literature review introducing the paradigm between entrepreneurship and social policy making. Then, it contextualizes the concept of SEE and the cases of study, Mexico and Chile. Following this, it describes the methodology used to conduct this empirical study. The empirical findings are then analyzed and discussed, highlighting their theoretical and practical contributions. This paper provides the final study contribution, limitations faced, and recommendations for future research venues in the Concluding Remarks section.

2. Literature Review

2.1. Entrepreneurship and the Social Policy Paradigm

For a new social policy paradigm to take shape, it must provide a general blueprint for promoting the wellbeing of society and individuals [3]. Since governments and markets have not effectively addressed social issues that have created undignified living conditions for millions of people, social entrepreneurs attempt to fill this gap [38]. Social entrepreneurship has gained increasing attention from academia and policymakers because of its impact on creating social and economic development [39]. It solves social problems in society and could also change the concept of social policy at the state and regional levels [40]. Social entrepreneurship could also be a framework within a social action context and describe an entrepreneurial ecosystem’s hybrid organization and structure [20,26,37]. It provides a catalytic leadership to raise social concern and rapid change by solving social issues and influencing social policy [41]. Hence, it can drive systemic social change [42]. It significantly influences social value, social capital, and the performance of the social enterprise [43].

Social entrepreneurs aim to generate new ideas to solve social problems and permanently change the system that caused these problems [33] and its institutions [44]. A new social paradigm is then identified when, on the one hand, there are increasing social issues, and social entrepreneurship seems a reliable source to solve them. Still, at the same time, social investments and support mechanisms are lacking because of the local conditions prescribed in the social policies. The solution could be found in having adequate social policies that properly shape the ecosystem to create and develop social enterprises, mainly where there are more significant social problems. In other words, the governments have the power to intervene in such economic processes as a dominant actor that can initiate divergent institutional change [45,46,47,48]. However, the traditional systems consider all entrepreneurial activities the same since, by definition, social entrepreneurship employs conventional entrepreneurship tools to pursue a social mission [49].

Social entrepreneurship implies and opens, as a consequence, space for various relationships with the government [3]. Its mission often involves reducing poverty, providing job opportunities to disadvantaged groups, granting access to products or services that were previously unavailable to people [50,51], and empowering people [2]; these objectives sometimes overlap with the SDGs. Social entrepreneurs seem to have a more profound sense of community that translates into having more inter-sectoral alliances [52]. Neumeyer and Santos [53] pointed out the need to understand under what conditions social mechanisms foster the ecosystem. Therefore, the ecosystem that supports social entrepreneurship is critical to helping these actors achieve sustainable business models and increase their social impact [9], given the established political–institutional systems to deliver local welfare services [12].

2.2. Social Entrepreneurial Ecosystems

In the entrepreneurship literature, the term ecosystem has several implications, depending on the outputs it is measuring; it can be used to refer to policies [54], regional clusters [55], and even national systems of entrepreneurship [56]. Due to its attractiveness and elasticity, the ecosystem concept has been used to explain various phenomena from various academic perspectives and under varying nouns, such as innovation, business, technology, platform, entrepreneurship, knowledge [57], and, more recently, sustainable ecosystems [23,26]. The main differences between them are the ecosystem outputs in terms of entrepreneurial activities and the units of analysis that are related to a thematic area, although they share the same characteristics of interdependent actors and factors as entrepreneurial ecosystem definitions do [58,59,60,61,62,63,64].

In Mexico, entrepreneurial diversity impacts the region’s sustainable development [15]; hence, it is crucial to analyze social entrepreneurship as an element of a larger ecosystem, resulting in a SEE concept that has many implications for achieving sustainable development. There is some evidence from Mexico that social entrepreneurs are navigating conventional entrepreneurial ecosystems; however, they have some degrees of separation from traditional entrepreneurs, and they need to operate in their customized ecosystem (or sub-ecosystem, which can overlap with the conventional) [16]. Thus, the literature on entrepreneurial ecosystems will be explored more with all the benefits of previous evidence and understanding, and it will be helpful to policymakers and researchers [61,65]. Ecosystem mapping is an essential strategic planning exercise that helps social entrepreneurs be more aware of their operating environment and its risks and opportunities [66]. Moreover, social enterprises benefit from favorable institutional conditions, especially public sector expenditure and regulatory quality [67].

Additionally, some empirical studies about entrepreneurial ecosystems have focused on creating value at the regional level [13,68,69,70,71], pointing to the need for a regional assessment of SEEs since they also create value, albeit social. Therefore, there is a latent need to understand SEEs better, but the problem is how to study them since the field is underexplored [72]. Therefore, this study examines SEEs from the macro level (countries) to the meso level (vulnerability regions) of two upper-middle-income countries (Mexico and Chile) because the sub-national level of regions is the most appropriate spatial level to identify and measure entrepreneurial ecosystems since the regional entrepreneurship literature provides striking evidence that entrepreneurship is primarily a regional (or local) event [73]. According to the World Bank, middle-income countries are defined as lower-middle-income economies—those with a GNI per capita between USD 1036 and USD 4045—and upper-middle-income economies—those with a GNI per capita between USD 4046 and USD 12,535 (2021). Middle-income countries are home to 75% of the world’s population and 62% of the world’s poor. At the same time, middle-income countries represent about one-third of global GDP and are major engines of global growth. Despite these definitions, both countries have several regions with different levels of economic development, separated geographically but connected culturally and socio-politically.

For social entrepreneurship, it is even more relevant to conduct a regional assessment of the ecosystem since different types of social entrepreneurship influence the types of social policies in which social entrepreneurship is embedded, and it depends on the geographic focus [74]. Since SEEs in upper-middle-income countries are essentially an underexplored subject, they present an opportunity to conduct research that fills this gap in the literature because social entrepreneurs are also gaining increasing attention as potential actors in addressing current social issues [16]. Another relevant fact is the political heritage of weak and corrupt governments and public sectors that encourage entrepreneurial solutions to social problems [75]. As Jenson [3] argued, governments are financing social interventions but working less on the design of policies. Therefore, social entrepreneurs must persevere in facing the inefficient institutional frameworks prevalent in emerging economies [76]. Moreover, there are differences (e.g., capitalization) across sectors in a particular region due to the degree of collaboration between the key actors of the ecosystem and the inequality of opportunities based on the characteristics of each industry [15], which indicate a heterogeneous phenomenon. Ecosystem heterogeneity affects success and explains why some ecosystems drive more than others [26].

2.3. The Mexican Context

In Mexico, 1% of the population receives 21% of the national income [77]. The Gini index of the per capita income is 0.50 [78]; income inequality is associated with more violence. Femicides and drug-related homicides in Mexico have increased in recent decades [1]; 71% of the population in urban areas believe that where they live is insecure [79]. Guerrero and Urbano [80] found that this institutional context of insecurity substantially affects the decision to implement social innovations in Mexico. Poverty is another issue: 52 million people (41.9% of the population) live in poverty, and 9.3 million live in extreme poverty [81]. These problems have prompted social entrepreneurs to find market-oriented solutions.

Mexico is one of the most dynamic countries in social entrepreneurship [82]. Mexico’s support system for social entrepreneurship is young but strong; it has specialized intermediaries and physical spaces supporting social entrepreneurs, although most are not more than five years old [83]. The cooperation between universities, public administrations, social enterprises, and NGOs is intense [84]. It works to create, develop, and consolidate entrepreneurial activity and entrepreneurship and innovation ecosystems [85]. However, Villegas-Mateos and Vázquez-Maguirre [16] found that the interactions and impacts of these participants, especially universities, often concentrate on low-vulnerability (LV) regions. There is a gap between the three major Mexican cities (Mexico City, Monterrey, and Guadalajara), where most of the actors concentrate, and the rest of the country [86]. Since most high-vulnerability (HV) regions lack a SEE supporting social ventures, some communities have created incipient local SEEs where a social or community-based enterprise is a central player that organizes and promotes the ecosystem [9]. The embeddedness of the economic relations in social relations in these regions facilitates this process [87]. However, these incipient SEEs remain local and do not integrate into a conventional entrepreneurial ecosystem, limiting their growth and impact [16]. Consequently, social enterprises are small entities with a limited impact that struggle to be sustainable [84]. Moreover, social entrepreneurs must adopt different legal frameworks that limit tax deductions [33] and potential access to resources and markets.

2.4. The Chilean Context

Chile has historically been an unequal country in terms of income, with a Gini index of 0.50 in 2017; it has the highest rate of income inequality among the first 50 countries with the highest human development in the world [1]. While 50% of poor households had an average net wealth of USD 5000, the wealthiest 10% had an average of USD 760,000, and the wealthiest 1% had USD 3,000,000 [78]. The 2017 United Nations Development Programme survey found that people are discontent with unequal access to healthcare (68%), education (67%), and respect and dignity (66%); social class and gender discrimination exacerbate inequalities [88]. These perceptions contrast with the fact that Chile has the lowest poverty rate in the region (10.7%), and it allocates the most resources per capita to social policies, which was USD 2387 in 2016 [78].

In the last decade, Chile has experienced a new wave of social ventures that are partially prompted by social movements [89]. This sector has faced significant crises; many social enterprises were forced to cease operations because of political repression under dictatorship rule [90]; then, the return to democracy in the 1990s allowed some of these ventures to be restored. The social enterprises in Chile have many different goals (environmental, political, social, and community). Still, they ultimately aim to build new social and labor relations that reduce inequalities and constitute an alternative to the capitalist economic system [89,91]. Although Chilean law has a legal framework for cooperatives, indigenous organizations, mutual societies, non-profit enterprises, and non-profit foundations, it lacks a specific legal framework for social enterprises [89]. Radrigán and Barría [92] argued that the SEE in Chile is characterized by a scarce integration of the actors, which may lead to fragmentation of the sector in the future. Additionally, Mancilla and Amorós [93] found that the entrepreneurship probability of Chileans who live in a secondary city is less, indicating that entrepreneurs may have different levels of institutional support in other regions.

3. Methodology

3.1. Data Description

The Global Entrepreneurship Monitor (GEM) conducts two surveys to collect data; ~50 countries participate. The datasets have been consistently used to develop empirical studies to compare the participating countries [94]. One of the surveys is the National Expert Survey (NES), which uses a standardized methodology to measure the Entrepreneurial Framework Conditions (EFCs). Using a standard instrument and method to obtain the data in different countries is an advantage of this survey because it allows comparisons between countries [95]. The authors found that some studies have used the EFCs to measure entrepreneurial opportunities in ecosystems of both countries in this study, Mexico [96] and Chile [97,98]. However, the EFCs primarily measure experts’ perspectives toward specific situations, which can be biased because they are based on their own experiences of traditional or social entrepreneurship. However, in 2015, the GEM teams around the globe made an extraordinary effort to collect data with their two surveys on a special topic report on social entrepreneurship [99]. The data about Mexico have been used before to assess the social policies and social ecosystem dynamism (SEEs) in the country at the regional level [16], and this study follows a similar method. It is relevant to the literature on entrepreneurial ecosystems since it follows the research on ecosystem dynamics, processes, territorial boundaries, and outcomes [26,61,63,100,101].

In 2015, Mexico, Chile, Germany, Spain, and England used a specific regional approach to collect the data. Each participating country that used the GEM’s methodology [95] in the data collection had only ~36 expert responses. Still, the countries that used the regional approach had at least 36 responses by the number of entities or regions participating. Mexico and Chile are the only Latin American countries that followed the regional data collection approach. In Mexico, the 2015 NES data were collected in 5 of the 32 entities in the country (Guanajuato, Jalisco, Puebla, Querétaro, and San Luis Potosí), resulting in 188 responses. In Chile, the data were collected in 7 of the 16 regions in the country (Antofagasta, Araucanía, Bío-Bío, Coquimbo, Metropolitana de Santiago, Tarapacá and Valparaíso), resulting in 276 responses (see

Table 1. Experts’ characteristics and numbers per country.

Entity	Mexico	Region	Chile
Guanajuato	38	Antofagasta	39
Jalisco	37	Araucanía	47
Puebla	38	Bío-Bío	36
Querétaro	39	Coquimbo	37
San Luis Potosí	36	Metropolitana	38
		Tarapacá	38
		Valparaíso	41
Total Mexico	188	Total Chile	276

). These datasets are appropriate for this study because the NES uses experts’ informed judgments about the status of social policies and social ecosystem dynamics in their entities and regions. This study is based on the GEM’s NES of 2015, using the most extensive available regional datasets of Latin American countries (Mexico and Chile) to analyze the 464 experts’ perceptions of these countries’ social entrepreneurial ecosystem conditions. In addition, this study follows a unique approach to group the regions of each country by the level of vulnerability, reducing the variance of the results across time. In fact, given the global socioeconomic crisis triggered in 2020 by the COVID-19 pandemic, UNESCO, WHO, and the IMF, among others, have recognized that the social problems increased even more in vulnerable regions, Latin America being the most affected. Consequently, the findings remain valid unless a significant economic development event occurs in these countries.

3.2. Sample Characteristics

The data comprise 464 interviews with experts, 188 from Mexico and 276 from Chile. The total sample was segmented into two sub-samples of HV and LV regions. Geographic, demographic, and economic indicators, such as the poverty indices of both countries, were considered in differentiating HV and LV regions (see

Table 2. Regional geography, demography, and economic indicators.

Country	Region	Male	Female	Total a/	% National GDP	% Poverty b/
Low-vulnerability regions
Chile	Antofagasta	315,014	292,520	607,534	9.53%	5.10%
Mexico	Guanajuato	2,832,687	3,032,090	5,864,777	3.90%	43.40%
Mexico	Jalisco	3,853,584	4,026,955	7,880,539	6.26%	28.40%
Chile	Metropolitana de Santiago	3,462,267	3,650,541	7,112,808	44.47%	5.40%
Mexico	Querétaro	995,355	1,048,496	2,043,851	2.07%	27.60%
Chile	Tarapacá	167,793	162,765	330,558	2.24%	6.40%
Chile	Valparaíso	880,215	935,687	1,815,902	7.32%	7.10%
Total		12,506,915	13,149,054	25,655,969
High-vulnerability regions
Chile	Araucanía	465,131	492,093	957,224	2.04%	17.20%
Chile	Bío-Bío	750,730	806,075	1,556,805	7.22%	12.30%
Chile	Coquimbo	368,774	388,812	757,586	2.91%	11.90%
Mexico	Puebla	2,949,444	3,233,876	6,183,320	3.05%	58.90%
Mexico	San Luis Potosí	1,321,029	1,402,743	2,723,772	1.92%	43.40%
Total		5,855,108	6,323,599	12,178,707

a/ Population and economic indicators by selected locations (Mexico, 2015; Chile, 2017). b/ Percentage of people in poverty (CASEN, 2017; CONEVAL, 2018).

). The differences between the regions were first tested in a study about Mexico using the same data, which is composed of five entities [16]. Once again, in this study, the sample comprises five entities, where two (San Luis Potosí and Puebla) are HV regions, and three (Jalisco, Querétaro, and Guanajuato) are LV regions. In Chile, the sample comprises seven regions: three (Araucanía, Bío-Bío, and Coquimbo) HV and four (Antofagasta, Metropolitana de Santiago, Tarapacá, and Valparaíso) LV regions. The HV regions in both countries are above their national average poverty rate, above 43.4 in Mexico [102] and above 8.6 in Chile [103]. Moreover, the HV regions contribute less to their national GDP. There are 194 responses from HV regions and 270 from LV regions in both Mexico and Chile. When their poverty rates are compared to their GDP per capita, which is USD 8902.8 in Mexico and USD 15,345.4 in Chile, Mexico has a higher vulnerability than Chile [104]. The characteristics of the sample can accurately test whether the vulnerability influences the perceived support mechanisms for social entrepreneurial activities in upper-middle-income countries. A detailed description of the total sample is provided in

Table 3. Total sample composition (N = 464).

Sample Characteristics		Total a/	% of Total b/	Mexico a/	% of Total b/	Chile a/	% of Total b/
Demographics	Average age	42.5		43.9		41.6
	Female	126	27.2%	49	26.1%	77	27.9%
	Male	338	72.8%	139	73.9%	199	72.1%
Educational attainment	Secondary	2	0.4%	0	0.0%	2	0.7%
	Vocational professional	8	1.7%	6	3.2%	2	0.7%
	University/college	182	39.2%	76	40.4%	106	38.4%
	MA, Ph.D.	270	58.2%	105	55.9%	165	59.8%
	Do not know	2	0.4%	1	0.5%	1	0.4%
Primary entrepreneurial framework conditions and expert specialization	Financial support	40	8.6%	21	11.2%	19	6.9%
	Government policies	47	10.1%	20	10.6%	27	9.8%
	Government programs	46	9.9%	20	10.6%	26	9.4%
	Education and training	55	11.9%	21	11.2%	34	12.3%
	R&D transfer	45	9.7%	22	11.7%	23	8.3%
	Commercial and professional infrastructure	46	9.9%	22	11.7%	24	8.7%
	Market openness	58	12.5%	22	11.7%	36	13.0%
	Access to physical infrastructure	41	8.8%	20	10.6%	21	7.6%
	Cultural and social norms	48	10.3%	20	10.6%	28	10.1%
	Data missing	38	8.2%	0	0.0%	38	13.8%
Expert specialization	Entrepreneur	329	70.9%	144	76.6%	185	67.0%
	Investor, financer, banker	72	15.5%	39	20.7%	33	12.0%
	Policymaker	181	39.0%	89	47.3%	92	33.3%
	Business and support service provider	287	61.9%	123	65.4%	164	59.4%
	Educator, teacher, entrepreneurship researcher	226	48.7%	88	46.8%	138	50.0%

a/ Valid cases for each variable. b/ Percentage based on total valid cases for each variable.

, and the sub-samples of the HV and LV regions are presented in

Table 4. Sub-sample composition (N = 464).

Sample Characteristics		Total a/	% of Total b/	HV a/	% of Total b/	LV a/	% of Total b/
Demographics	Average age	42.5		41.2		43.5
	Female	126	27.2%	58	29.9%	68	25.2%
	Male	338	72.8%	136	70.1%	202	74.8%
Educational attainment	Secondary	2	0.4%	1	0.5%	1	0.4%
	Vocational professional	8	1.7%	6	3.1%	2	0.7%
	University/college	182	39.2%	80	41.2%	102	37.8%
	MA, Ph.D.	270	58.2%	106	54.6%	164	60.7%
	Do not know	2	0.4%	1	0.5%	1	0.4%
Primary entrepreneurial framework conditions Expert specialization	Financial support	40	8.6%	19	9.8%	21	7.8%
	Government policies	47	10.1%	23	11.9%	24	8.9%
	Government programs	46	9.9%	22	11.3%	24	8.9%
	Education and training	55	11.9%	27	13.9%	28	10.4%
	R&D transfer	45	9.7%	18	9.3%	27	10.0%
	Commercial and professional infrastructure	46	9.9%	19	9.8%	27	10.0%
	Market openness	58	12.5%	23	11.9%	35	13.0%
	Access to physical infrastructure	41	8.8%	20	10.3%	21	7.8%
	Cultural and social norms	48	10.3%	23	11.9%	25	9.3%
	Data missing	38	8.2%	0	0.0%	38	14.1%
Expert specialization	Entrepreneur	329	70.9%	143	73.7%	186	68.9%
	Investor, financer, banker	72	15.5%	16	8.2%	56	20.7%
	Policymaker	181	39.0%	87	44.8%	94	34.8%
	Business and support services provider	287	61.9%	121	62.4%	166	61.5%
	Educator, teacher, entrepreneurship researcher	226	48.7%	86	44.3%	140	51.9%

HV = high-vulnerability experts; LV = low-vulnerability experts. a/ Valid cases for each variable. b/ Percentage based on total valid cases for each variable.

. Pearson’s chi-square tests were carried out to evaluate the similarities of the samples, indicating that they are not significantly different.

3.3. Measures

Social entrepreneurship mechanisms in the GEM’s NES were measured with eight additional questions in a new section at the end of the 2015 survey. The questions were measured with a nine-point Likert scale (where 1 = completely ‘false’ and 9 = completely ‘true’). The authors conducted an exploratory factor analysis and found that the eigenvalues of two components are greater than 1. The first four questions measure social policies, and the other four measure social ecosystem dynamism. Each set of four items was selected to check its internal consistency using Cronbach’s alpha test because it indicates the extent to which a set of questions or items from a survey measures a single unidimensional latent construct [105]. The results of these analyses are presented in

Table 5. Scale reliability.

Scales	Number of Items	Cronbach’s Alpha
Social policies	4	0.513
Social ecosystem dynamism	4	0.673

. Nunnally [106] recommended a level of 0.7 or above for the alpha coefficient; in this study, both scales are below 0.7, where social policies are 0.513, and social ecosystem dynamism is 0.673. However, there is also a generally accepted rule that an alpha coefficient of 0.6–0.7 indicates an acceptable level of reliability [107], providing evidence of sufficient reliability, which is consistent with the cross-national use of the NES. Although social policies’ Cronbach’s alpha of 0.513 is at a questionable level, the authors decided to use it based on experts’ judgments. Previous studies that used this type of data about the region [16,96,97,98] agree that one limitation is the scale’s reliability, but it can be improved by increasing the sample size, which this study does. With this Cronbach’s alpha test’s results, procedures like the principal component analysis (PCA) or factor analysis can be used as variable reduction procedures to reclassify the set of questions in the social entrepreneurship section of the NES into two new variables (social policies and social ecosystem dynamism) to analyze them as described in the next section.

3.4. Method

Once the reliability tests were conducted and the results were statistically appropriate, the authors used the PCA to reclassify each set of the four social entrepreneurship-related questions into two new summarized variables. The PCA is a well-known, helpful, powerful, multivariate tool for analyzing complex data and reducing dimensionality using a linear combination of optimally weighted observed variables [108,109,110,111,112]. The method generated two new summarized constructs (social policies and social ecosystem dynamism) that contain most of the variation in the data [113], and they can be used to compare the differences in the countries (Mexico and Chile) and regions (HV and LV). Amorós et al. [97] used this method to analyze entrepreneurial opportunities in core and peripheral regions in Chile, and Villegas Mateos and Amorós [96] used the same approach to analyze central and non-central EEs in Mexico with all the EFCs of the GEM’s NES. Furthermore, in 2020, Villegas-Mateos and Vázquez-Maguirre [16] examined the SEEs in LV and HV regions. Hence, the PCA is preferred over the factor analysis because of its validation in similar empirical studies about upper-middle-income countries.

However, the Bartlett and Kaiser–Meyer–Olkin (KMO) tests were conducted in a prior internal consistency validation process of the survey. The KMO statistic is 0.591 for social policies and 0.707 for social ecosystem dynamism, which is acceptable since it is above 0.5, indicating that the PCA is appropriate for the sample [114]. Additionally, both have a high significance level (p-value = 0.000) from the Bartlett test [115]. The comparisons between countries and regions were possible after the PCA was conducted (see

Table 6. Description of the principal component analysis.

Social Policies
Statement					Communalities extraction	Component matrix
In my region, people who live in poverty cannot rely on the government or civil society organizations					0.273	0.523
In my region, you will find many business that provide people with basic needs that are covered by governments and civil society organizations in other countries					0.430	0.656
In my region, social, environmental and community problems are generally solved more effectively by businesses than by the government and civil society organizations					0.485	0.696
In my region, entrepreneurs’ associations/groups challenge existing regulations that negatively impact particular groups in society or the environment					0.447	0.669
Social ecosystem dynamism
Statement					Communalities extraction	Component matrix
In my region, the government is able to bring together potential entrepreneurs, businesses and civil society organizations around specific social, environmental or community projects					0.531	0.729
In my region, consumers are putting pressure on businesses to address social and environmental needs					0.345	0.588
In my region, there are sufficient private and public funds available for new and growing firms that aim at solving social and environmental problems					0.661	0.813
In my region, there is a lot of media attention for new and growing firms that combine profits with positive social and environmental impact					0.520	0.721
Total variance explained
Component	Initial eigenvalues			Extraction sums of squared loadings
	Total	% of variance	Cumulative %	Total	% of variance	Cumulative %
1	1.635	40.878%	40.878%	1.635	40.878%	40.878%
2	1.035	25.863%	66.741%
3	0.693	17.331%	84.072%
4	0.637	15.928%	100.000%
Component	Initial eigenvalues			Extraction sums of squared loadings
	Total	% of variance	Cumulative %	Total	% of variance	Cumulative %
1	2.057	51.420%	51.420%	2.057	51.420%	51.420%
2	0.789	19.722%	71.142%
3	0.679	16.963%	88.105%
4	0.476	11.895%	100.000%

). Thus, the authors chose the best test for mean comparison, which depended directly on the data distribution. Therefore, normality tests were conducted (Kolmogorov–Smirnov and Shapiro–Wilk). These tests reveal that neither social policies nor social ecosystem dynamism are normally distributed in the groups. Thus, the Mann–Whitney U non-parametric test for mean comparison was selected as the most appropriate method to compare both groups because it is considerably more efficient and robust than the t-test when sample distributions are not normal [116].

4. Results and Discussion

The Mann–Whitney U tests are reported in

Table 7. Mann–Whitney U test results.

Scale	Group	Valid Cases	Mean	Standard Deviation	Mean Ranges	Mann–Whitney U	Z	p-Value
Social policies	Mexico	165	4.99	1.62	228.57	17,066.0	−2.917 **	0.004
	Chile	249	4.60	1.40	193.54
	HV	171	4.61	1.48	196.11	18,829.0	−1.625	0.104
	LV	243	4.85	1.51	215.51
Social ecosystem dynamism	Mexico	175	4.17	1.79	217.71	22,625.5	−0.029	0.977
	Chile	259	4.14	1.32	217.36
	HV	183	4.32	1.58	232.05	20,303.0	−2.064 *	0.039
	LV	251	4.04	1.48	206.89

* p ˂ 0.05; ** p ˂ 0.01 (two tailed).

. At the macro level, the perception of experts in Chile about social policies for social entrepreneurship is higher than that of experts in Mexico, with a high significance level (z = −2.917, p = 0.004). One possible reason is the relative success of Chile, as compared with Mexico, in addressing poverty and funding social problems. Chile allocated the most resources per capita for social policies in Latin America (USD 2387 in 2016) and decreased its poverty rate to 10.7%. By contrast, Mexico allocated less than USD 1000, and its poverty rate was 41.9% in 2016 [78]. Moreover, the Mexican government has specialized entities and policies for entrepreneurs and social issues, but different actors are not working together. They feel reluctant to work with the public sector [83], which could explain the low level of the experts’ perception of social policies.

At the meso level, social policies are perceived as more favorable in HV regions than in LV regions, but the results are not statistically significant. However, regarding social ecosystem dynamism, the results are significantly better in HV regions (z = −2.064, p = 0.039). This result is inconsistent with previous studies [86,93], which suggest that the level of support for social entrepreneurship in LV regions is perceived as more favorable than that in HV regions. In Mexico, Saiz-Álvarez and Rodríguez-Aceves [84] found intense cooperation that promotes a dynamic social ecosystem between universities, public administrations, social enterprises, and NGOs. However, these actors are usually located in urban areas with LV rates. Hence, their impact is generally concentrated in these regions.

Moreover, the HV regions are often located in rural areas with difficult accessibility, and reaching these areas may be costly for some actors in terms of budget and time. Thus, it is reasonable that some actors set their operations in the country’s main cities, where they can be easily accessible to partners, resources, and favorable institutional support [67]. One possible reason for this outcome is the presence of other types of networks in HV regions. Vázquez-Maguirre [9,117] documented the creation of local entrepreneurial micro-ecosystems in HV regions in Mexico. Since these Mexican regions are usually more isolated, local actors mobilize resources that are typically embedded in a structure of social relations to create value that benefits entrepreneurs and the entire region [87]. Because some actors (e.g., universities, financial institutions, incubators, and accelerators) prefer to operate in cities that provide a high quality of life, HV regions must create local ecosystems to address their social issues. Fehrer et al. [36] also studied these local micro-ecosystems that target specific social problems and suggested that they are a promising research topic. These micro-ecosystems are not usually connected to regional or national ecosystems. Still, they are highly effective in addressing social issues [9,117], so experts in the survey may have perceived these micro-ecosystems as more dynamic than those in LV regions.

SEEs are primarily needed in peripheral regions, which traditionally have higher levels of vulnerability [16]. The HV regions have energetic and resourceful entrepreneurs who manage to make a lot out of very little to improve their living conditions [118]. SEEs can take advantage of these people’s entrepreneurial spirit and help them start or grow their businesses and grant them access to transformative services. However, it remains challenging to enable an environment for social entrepreneurship through government social policy worldwide [19]. Thus, the results may imply that regions with high poverty rates and lower contribution to GDP (see Table 2) presumably have a more dynamic social ecosystem that probably targets specific local problems. In this sense, Vázquez-Maguirre [10] documented cases of social enterprises in HV regions of Mexico, the ecosystem they created, and how they targeted specific SDGs related to good health and wellbeing (SDG 3), decent work (SDG 8), reduced inequalities (SDG 10), and sustainable communities (SDG 11). Although they are located in HV regions, they were able to build a regional ecosystem that worked to punctually address their social, economic, and environmental issues [10].

5. Concluding Remarks

Since the world faces rising inequalities and more people live in extreme poverty, social entrepreneurs in entrepreneurial ecosystems actively seek sustainable solutions to these and other issues. This study explores how a region’s degree of vulnerability influences the perceived level of social policies and social ecosystem dynamism in Mexico and Chile. The results suggest that the level of vulnerability positively impacts the social entrepreneurship mechanisms that underlie the local ecosystems. Therefore, regions with high poverty rates and low contribution to the national GDP presumably have higher social policies and social ecosystem dynamism. This can be contradictory between theory and practice, with policymakers working in theory to implement the national economic development agenda, but in practice, the more vulnerable regions (sometimes more populated) benefit from the social policy programs. A sound SEE in HV regions can take advantage of people’s entrepreneurial spirit and help social entrepreneurs design permanent solutions that will create new equilibriums and structural changes for millions who are currently denied a decent life. Moreover, SEEs may constitute a feasible strategy to address urgent social issues and contribute to achieving the SDGs while multiplying the economic value. The implications are practical and may require that policymakers and practitioners set better strategies to create and develop social enterprises in upper-middle-income countries to reach sustainable economic development with mechanisms that generate returns on the impact investment.

Future research may focus on how SEEs evolve in HV regions and the microdynamics that help generate community growth and consolidation. Moreover, the entrepreneurial micro-ecosystem dynamics that develop in local contexts and target specific social issues need further examination, especially in HV regions and upper-middle-income countries. Another avenue is to explore the different actors and activities essential for social entrepreneurs in both LV and HV regions to consolidate their social enterprises. In the entrepreneurial ecosystem literature, this research is one of the first attempts to study an underexplored field that requires more examination at different levels, especially at the meso and micro levels. The contribution to the academic field of entrepreneurial ecosystems discusses the differences with other emerging research lines, such as sustainable entrepreneurial ecosystems. A limitation of this research is that it does not ensure that the measure of social policies and social ecosystem dynamism for social entrepreneurship is the best and unique way to fill the gap in the literature. The data collection can be a problem for researchers because of the level of analysis required.