Factors that explain entrepreneurship

The term entrepreneur is related to the resolution and determination to start a business amid danger, difficulty, and uncertainty. It originally referred to adventurers, then contractors, until it shifted to meanings related to business activity and function (^{Arango, 2011}; ^{Arango, 2017}). The entrepreneur seeks business opportunities, takes risks, and has intelligence, optimism, creativity, and persistence in achieving the goals they set for themselves (^{Brunet & Alarcón, 2004}; ^{Corrêa et al., 2017}; ^{Terapuez & Botero, 2007}). The entrepreneur has innate abilities (trait theory), feels high job satisfaction (behavioral theory), is capable of multitasking (role theory), and can interact in diverse situations efficiently (situational theory) (^{Escandon & Hurtado, 2016}).

In order to start a business, self-knowledge as an entrepreneur is necessary (^{Valenciano & Uribe, 2009}; ^{Ynzunza et al., 2020}). Knowing one's limitations and strengths helps to identify the right conditions to undertake a new business (^{Pazmiño et al., 2018}; ^{Plaza, 2015}; ^{Silveira et al., 2015}). In a similar sense, the experience acquired enables the entrepreneur to manage their business better and increase the probability of achieving business success, which is also related to the ability to match business strategies with the situation of the sector and the needs of the clients (^{Escandon & Hurtado, 2014}; ^{Escandon & Hurtado, 2016}; ^{Franco & Urbano, 2010}). Nevertheless, education and entrepreneurial ability are not necessary conditions to initiate a new business (^{Diez et al., 2021}).

Factors explaining the company

Companies exist because markets are imperfect and because through them the means of production are combined to produce goods and provide services whose prices are determined by market fluctuations (^{Brunet & Alarcón, 2004}; ^{Coase, 1937}; ^{Urbano et al., 2007}). Companies also exist because of contracts, transactions and exchanges, which enable the transfer of assets between different stakeholders (^{Urbano et al., 2007}; ^{Williamson, 1985}). The company can be the result of rational decisions (rational choice theory) that drive the search for economic benefit (expected profit theory), the result of the occurrence of a negative critical event (marginalization theory), or the result of the ability to integrate and leverage controlled resources optimally and strategically (resources and capabilities theory) (^{Brunet & Alarcón, 2004}; ^{Franco & Urbano, 2010}; ^{Martínez & Martínez, 2008}; ^{Morales & Segoviano, 2016}).

Companies differ despite carrying out similar activities and competing in similar markets (Cuervo, 2004). Consolidated companies adapt easily to the environment, promote cooperation, satisfy needs, contribute to local development, do what is useful and necessary, spread solidarity, innovate, develop new products and services, capitalize, and know how to take advantage of environmental opportunities (^{López & Calderón, 2006}; ^{Sanchis, 2001}). Human resource management, task assignment, training and education, teamwork, motivation, a trained entrepreneur, and staff that complements their deficiencies are aspects that favorably affect the growth and consolidation of the company (^{Carmona et al., 2015}; ^{López & Calderón, 2006}; ^{Valenciano & Uribe, 2009}; ^{Vílchez et al., 2019}).

Factors that explain the environment

The environment provides cultural factors and values that impact the entrepreneur and the company (^{Alvarez et al., 2010}; ^{Camino & Aguilar, 2017}). The behavior of the entrepreneur and the company is defined by regulations and rules (institutional economic theory) that through selectivity condition their permanence (ecological theory of population) in an environment whose interconnection of its agents (network theory) determines their consolidation to the extent that their actions are perceived as credible (role theory) (^{Brunet & Alarcón, 2004}). Likewise, the influence of the environment is linked to local production systems that are also related to Marshallian industrial districts, in which it is suggested that there are sectors that are selectively more productive than others, either for geographical, social, cultural, or economic reasons (Boix & Galletto, 2005; ^{Brunet & Alarcón, 2004}; Climent, 1997).

Companies distrust each other and try to hide their strengths, weaknesses, and difficulties; they are very closed and selective in their commercial relations (^{López & Calderón, 2006}). The aid that companies receive from the environment can contribute to their consolidation and improve their competitiveness, innovation, and development (^{Cardona et al., 2008}; ^{García et al., 2017}; ^{Plaza, 2015}; ^{Valenciano & Uribe, 2009}). The environment influences entrepreneurial activity (^{Chaves et al., 2018}) and provides aids such as business advice, partnership, and funding sources that are conditional on the consolidation of companies (^{Bada et al., 2017}; ^{Sanchis, 2001}; ^{Valenciano & Uribe, 2009}). Thus, the cooperative approach, which is based on support networks, collectivity, and associativity, is a key factor in the success of companies (^{Di Masso et al., 2021}; ^{Kasparian & Rebón, 2020}; ^{Mera et al., 2018}). Moreover, this approach could be applied to all types of companies regardless of their legal form, whether for-profit or not-for-profit.

Factors explaining consolidation

Business consolidation is related to several factors that determine it (^{Plaza, 2015}). According to the Association of Young Entrepreneurs (AJE) (Spanish: Asociación de Jóvenes Empresarios), business consolidation creates the conditions for the company to endure over time under profitable conditions (^{AJE, 2015}). Consolidation is a challenge for new companies and a characteristic of success for those that have endured in the markets in which they operate (^{Anaya, 2014}). Consolidation is determined by economic, commercial, technological, social, and cultural aspects (external factors), as well as by commitment, quality, service, loyalty, compliance, and dedication to work (internal factors) (^{López & Calderón, 2006}). Nonetheless, the company's consolidation is framed by internal tensions generated by the demand for productivity and competitiveness to achieve business success (^{Bastida et al., 2020}; ^{Figari, 2019}).

Business consolidation follows a gradual process of planning, organization, and execution of technical, human, and financial resources toward an advanced state of innovation, which largely depends on the degree of productivity, competitiveness, and entrepreneurship achieved (^{AJE, 2015}; ^{Melo et al., 2021}; ^{Valenciano & Uribe, 2009}). Consolidation is reflected in the value of assets, employment generated, organizational status, business profit, and decreased risk (^{Jurado, 2018}; ^{Sanchis, 2001}). Strengthening organizational resources and capabilities contributes to the consolidation of the business project (^{Acosta, 2013}; ^{Araya et al., 2017}). Likewise, sectoral support, financing, taxation, research, innovation, organizational culture, administrative procedures, and the qualifications of the entrepreneur are essential factors in any business consolidation process (^{Camino & Aguilar, 2017}; ^{Foncubierta et al., 2020}; ^{Plaza, 2015}; ^{Texis et al., 2016}).

Lack of planning, management, quality, training, and technical capabilities, as well as low productivity and competitiveness, weakness in obtaining suppliers, failures in defining a market niche, not having facilities, informality in organizational processes, and lack of knowledge of the sector, are factors that affect the consolidation of the companies (^{AJE, 2015}; ^{Durán & San Martin, 2013}; ^{Rodríguez & Dussán, 2018}; ^{Texis et al. 2016}). Likewise, there is a positive relation between business consolidation and the professional option of creating a new company by keeping adequate accounting and finances in line with the requirements of recording the economic transactions of the new business (^{Álvarez et al., 2011}; ^{Ibarra et al., 2017}).

Consequently, business consolidation depends more on integrated factors of the entrepreneur, the company, and the environment than on factors viewed individually (Table 1).

Research hypothesis

The research hypotheses are correlation and relation hypotheses (^{Hernández et al., 2014}). In order to establish these hypotheses, Entrepreneur Factors (FER), Company Factors (FEM), and Environment Factors (FET) were defined as independent variables, and Business Consolidation Factors (CEM) as the dependent variable (Table 2).

Subject of research

This paper aims to identify the factors that determine business consolidation based on integrating factors associated with the entrepreneur, the company, and the environment. Statistical analysis could determine this integration based on the correlation and relation between the variables under study.

Observation instrument

The information was obtained through a questionnaire structured in nine dimensions of Demographic Factors (DF), eight Entrepreneur Factors (FER), seven Company Factors (FEM), eight Environment Factors (FET), and five Business Consolidation Factors (CEM). The evaluation scale was a Likert scale from 1 to 4, where 1: strongly disagree, 2: disagree, 3: agree, and 4: strongly agree (Table 3).

Theoretical model

Based on the variables and their dimensions, the theoretical model of correlations and relations was defined to respond to the research hypotheses (Figure 1).

Source: created by the author

Figure 1 Theoretical model of correlations and relations 

Sample selection

The target population was large, medium, small, and micro enterprises in the Latin American context. The sample was selected by simple random sampling. With a finite population, 95% confidence level and maximum variance, the sampling error was 6.32%. The questionnaire was e-mailed to 1 800 businesspeople and managers of commercial, industrial, service, and agricultural companies. The data collection time was eight months between 2018 and 2019, with a response rate of 11.78%, corresponding to 212 respondents.

Descriptive analysis

The largest share by country is in Colombia (91.9%), followed by Mexico (2.8%), Bolivia, and Chile, with shares close to 2%, and Spain and Panama, with less than 1% shares. Regarding corporate purpose, the greatest participation was in service companies (66.4%), followed by commercial companies (17.5%). Regarding the type of company, micro companies have the largest share (38.9%), followed by small (26.5%), while medium and large companies have shares slightly above 17% (Table 4).

Statistical model

The statistical model is a structural model constructed from exploratory and confirmatory factor analysis. In this model, the FER, FEM and FET factors are the independent and exogenous variables, and the CEM factors are the dependent and endogenous variables. There are three correlation hypotheses among FER, FEM, and FET: H₁, H_{2, and} H₃; among FER, FEM, FET, and CEM, there are three relation hypotheses: H₄, H₅, and H₆ (Figure 2).

Source: created by the author

Figure 2 Statistical model 

Normality and homogeneity

The normality of the variables was determined using the Kolmogorov-Smirnov (K-S) test, which compares the theoretical distribution function with the empirical distribution (^{Pedrosa et al., 2014}). A p-value ≤ 0.05 indicates that the variables do not meet the normality condition. The K-S test showed that the p-value of FER, FEM, and FET is less than 0.05, while the p-value of CEM is greater than 0.05. The homogeneity of the variables was determined using Levene's test, which establishes whether the variances of the observed groups are statistically equal. A p-value≤0.05 indicates that the variables do not meet the homogeneity condition. Levene's test showed that the p-value of FER, FEM, FET, and CEM is greater than 0.05, i.e., their variances are homogeneous (Table 5).

Considering that at least one of the criteria for parametric tests is not met (^{Berlanga & Rubio, 2012}; ^{Rubio & Berlanga, 2012}), the statistical tests performed were nonparametric, which is likewise corroborated by the descriptive results of skewness and kurtosis (Table 6).

Consistency and reliability

Sample consistency was determined using the Kruskal-Wallis (K-W) test, which ascertains whether the sample comes from identical or different populations. A p-value≤0.05 indicates that the sample comes from different populations. The K-W test showed that for FER and FEM, the sample comes from identical populations, while for FET and CEM, the sample comes from different populations. Meanwhile, the reliability of the observation instrument was determined by Cronbach's Alpha, which indicates the degree to which the different items are consistent and can be used to measure the same magnitude (^{Cupani, 2012}). Values close to 1 indicate that the instrument used is reliable. Cronbach's alpha showed that FER, FEM, FET, and CEM have values above 0.80, indicating that the instrument is reliable, the items that comprise it are consistent, and the data obtained can be used in statistical analyses (Table 7).

Construct validity

Construct validity enables the statistical validation of items grouped into factors and is determined by convergent and discriminant validity (^{Hayton et a., 2004}). Convergent validity is obtained through the Expected Cross Validation (ECV) and Composite Reliability (CF) indices. The closer the ECV and CF are to 1, the higher the correlation between the items. The ECV of the FER, FEM, FET, and CEM items indicated an acceptable correlation, while the CF showed that the FER and FEM items were highly reliable and that the FET and CEM items were reliable. Discriminant validity is measured with the Mean-Variance Extracted (MVE) index, which measures the theoretical differences between the study variables, which are expected to have lower correlations on a scale. The VME showed that the theoretical differences between FER, FEM, FET, and CEM are acceptable (Table 8).

Exploratory factor analysis

Since the original data did not satisfy the normality condition, the Exploratory Factor Analysis (EFA) was performed using nonparametric statistical tests, with an unweighted least squares extraction method, varimax rotation method with Kaiser normalization, and factor loading level greater than 0.50. Values with factor loadings lower than 0.50 were not considered for the definition of the structural model.

KMO index and Barlett's test

The Kaiser-Meyer-Olkin index (KMO) compares the correlations of the variables and their partial correlations. Only with values close to l can factor analysis be performed reliably. Barlett's test of sphericity establishes whether the correlation matrix is an identity matrix and, if so, it would not be appropriate to carry out the factor analysis. A p-value≤0.05 indicates that the matrix is not of identity and is appropriate to perform factor analysis (^{De la Fuente, 1999}). The KMO index showed that the sample adequacy of FER, FEM, and CEM is good, and that of FET is acceptable. Barlett's test showed that the correlation matrix is not identical and that it is appropriate to perform the factor analysis (Table 9).

Rotated factor matrix

The varimax rotation for unweighted least squares made it possible to group the items by factorial homogeneity and eliminate those with factor loadings lower than 0.50 (^{De la Fuente, 1999}). 12 of the 22 FER items were grouped into five factors, 12 of the 22 FEM items into five factors, 13 of the 23 FET items into five factors, and 11 of the 23 CEM items into three factors (Table 10).

The above items are those that statistically best explain the research phenomenon and are considered to estimate the structural model through Confirmatory Factor Analysis (CFA).

Confirmatory factor analysis

Confirmatory Factor Analysis (CFA) consists of estimating the statistical parameters that determine the correlations and relations between the structural model's variables, constructs, and indicators. The CFA allows the deletion of other items to seek a better model fit while enabling the statistical parameters to be contrasted for the corroboration or rejection of the research hypotheses (^{Escobedo et al., 2016}; ^{García, 2011}; ^{Mavrou, 2015}).

Structural model

The structural model integrates the statistical parameters of the correlations and relations between the independent and dependent variables in the same graph and essentially deals with testing the research hypotheses. Structural models are regression equations with measurement errors corresponding to unexplained variances between the independent and dependent variables (^{Cupani, 2012}; ^{Garcia, 2011}; ^{Lara, 2014}). Respecification allows the addition or removal of parameters according to higher factor loadings or lower measurement errors seeking an acceptable model fit (^{Cupani, 2012}; ^{Escobedo et al., 2016}).

Using the structural model, it was possible to determine a correlation of 0.69 and a mutual influence of 48% between FER and FEM. FER and FET have a correlation of 0.36 and a mutual influence of 13%. Furthermore, between FEM and FET there is a correlation of 0.90 and a mutual influence of 81%. On the other hand, the relation between FER and CEM is 0.32, between FEM and CEM is 0.32, and between FET and CEM is 0.30, i.e., an influence of 10%, 10%, and 9%, respectively, showing that CEM is explained proportionally by FER, FEM, and FET (Figure 3).

Source: created by the author based on AMOS 24 statistics

Figure 3 Structural model of FER, FEM, FET, and CEM 

Model identification

The model identification is defined in degrees of freedom (g). If g<0, there is no identification. If g=0 there is identification, and if g>0 there is over-identification (^{Escobedo et al., 2016}; ^{Lara, 2014}). The proposed model is expected to be over-identified (g>0). It presents multiple solutions because the data matrix has more information than the number of parameters to be estimated. The more degrees of freedom (g) the model has, the more parsimonious it is and, therefore, the better the fit of the data and the correlations and relations between the independent and dependent variables can be tested (^{Cupani, 2012}).

The calculations showed that the data present g>0 values, i.e., that the statistical models are over-identified, i.e., that they present multiple possible solutions and that the correlations and relations between the variables can be verified (Table 11).

Model fit

Model fit is determined based on goodness of fit indices, the calculation of which makes it possible to evaluate the extent to which the model reproduces the correlations and relations between the variables that define it (^{Escobedo et al., 2016}; ^{Lara, 2014}). The calculated goodness of fit indices were the root mean square error index (RMR), goodness of fit index (GFI), adjusted goodness of fit index (AGFI), parsimony goodness of fit index (PGFI), normed goodness of fit index (NFI), relative fit index (RFI), parsimony relation (PRATIO), and parsimony normed fit index (PNFI). The calculations show that the statistical models acceptably specify the correlations and relations between the independent and dependent variables (Table 12).

Statistical testing of hypotheses

Structural models allow the testing of research hypotheses (^{Escobedo et al., 2016}). Likewise, to corroborate the working hypotheses, the p found can be compared with the statistical significance defined by p-value ≤ 0.05. If the p-value found is less than the statistical significance, the null hypothesis is rejected; otherwise, it is accepted (^{Leenen, 2012}). The present research, which was based on the structural equation model, takes root mean square error ratio (RMR) as the test parameter (Table 12), which is less than the statistical significance and shows an acceptable fit of the model, thus enabling acceptance of the working hypotheses, which in turn is reflected in the correlations, relations, and influences shown by the structural model. From the results obtained, it is corroborated that there are positive correlations and relations between the variables under study (Table 13).

Business consolidation model

Based on the statistical results, a business consolidation model that relates the factors of the entrepreneur, the company, and the environment is proposed. The model is intended as a graphic reference to help simplify the complex relations between the factors that determine business consolidation and is not intended to be a definitive model (Figure 4).

Source: created by the author

Figure 4 Business consolidation model 

The above model considers that risk, perseverance, leadership, identity, and decisions are the factors that define the entrepreneur and have the greatest influence on the consolidation process of a company. Strategy, innovation, personnel, market, and profitability define the market and have the greatest influence on the consolidation process of a company. Likewise, financing, competition, networks, growth, and contacts are the factors that define the environment and also have the greatest influence on the consolidation of a company. Institutional, social, and environmental factors define such consolidation. In addition, there is an economic and social environment since there are factors associated with these dimensions that, although beyond the company's internal control, significantly influence its birth, survival, and consolidation. The statistical results of correlation and relation between the factors that were reduced through factor analysis reflect the fact that the model applies to any company, regardless of its corporate purpose or size. That is, not all companies need to be large to be consolidated companies in the markets in which they operate. Micro, small and medium-sized companies are consolidated when the factors represented in this model come together.

Limitations and implications

One limitation was collecting information from the surveyed population (entrepreneurs and company managers). Although a wide dissemination of the surveys was made with repetitive mailings and various time lapses between 2018 and 2019, out of 1800 surveys sent out, 212 were received. Although a larger response rate was expected, the sample collected is considered acceptable for factor analysis in statistical distributions that do not meet the normality condition. This situation may be due to a growing reluctance among entrepreneurs in Latin American countries to respond to surveys via e-mail, an aspect beyond the research's control. Although the research was intended to cover several Latin American countries, Colombia and Mexico were the countries that contributed the most, so the results are more influenced by businesspeople from these two countries.

A second limitation was the reduction through factor analysis of the original factors to a smaller number seeking a better model fit, which, although limited compared to a wider range, are considered sufficient to determine correlations and statistical relations with the factors that determine business consolidation.

A third limitation could be the Covid-19 effect on business consolidation since the data collection was before the pandemic², and it is evident that business dynamics were substantially affected by this phenomenon that has influenced social, economic, and business aspects. Nevertheless, it is considered that the research results are still valid, and future research should statistically analyze the correlation and relation of this type of phenomenon in the consolidation of companies.

The implications are that the entrepreneur, the company, and the environment are a unit that determines the durability of the companies over time. A possible subject of future studies could be to test the proposed model in other countries, sectors, and economic activities.