The effect of governance on entrepreneurship: from all income economies perspective

Abstract

The purpose of this study is to analyze the effect of governance indicators on Entrepreneurship. Explanatory research design with Pearson correlation and multiple linear regression models were applied. Five-year World Bank data (2014–2018) of 126 countries from all economic development levels were used. Worldwide governance indicators considered are voice and accountability, political stability, government effectiveness, regulatory quality, rule of law, and corruption control. Gross net income was taken as a control variable. To measure entrepreneurship, the number of formally registered limited liability businesses as a percentage of the working-age population, was used. To make highly skewed time series data of dependent variable (entrepreneurship) closer to normal, logarithmic transformation was made and heteroscedasticity of residuals was checked. The finding of Pearson correlation shows that there are moderate to strong significant correlations (0.466 ≤ r ≤ 0.806, p < 0.01) between predictors and the outcome variable. Predictor variables have moderate to very strong correlation among each other (0.506 ≤ r ≤ 0.956, p < 0.01). Regression analysis was computed after two highly collinear variables were dropped from the model using the variance inflation factor (VIF) test. The study found that the remaining four independent variables and the control variable predict 71.5% of the variance in the outcome variable. Except for voice and accountability, all predictors have their own statistically significant influence on entrepreneurship. Thus, working on each predictor up to the standard application can bring incremental changes in new business formation and entry. The researchers believe that this study is of significant interest to policymakers, program developers, entrepreneurs, analysis, and supporters, since it provides useful insight on how governance indicators influence entrepreneurship.

Introduction

Entrepreneurship has long been the concern of all countries regardless of their economic development level. It has been highly concerned to minimize their overstraining challenges of unemployment, poverty, and social unrest for all low-income and some medium-income countries. It is also the focus of attention in high-income level countries to make their development consistent. Politicians, policymakers, and individuals feel concerned not only about supplying potential entrepreneurs into their economy but also assuring sustainable growth of the existing one. This progressively growing interest in entrepreneurship as a means of problem solving and development has also had an influence on the research world (Audretsch, 2012). A large number of entrepreneurship studies on different entrepreneurship issues have been conducted at different times. The issues are largely characterized as the essence of entrepreneurship (Baumol & Schilling, 2008; Cieslik, 2017; Stevenson & Jarillo, 1990), its economic and social contribution (Demir-Uslu & Kedikli, 2019; Meyer & de Jongh, 2018; Parker, 2009; Valliere & Peterson, 2009), and entrepreneurial environment (Mason & Brown, 2014; Zamberi Ahmad & Xavier, 2012). Thus, updating the stated research dimensions paves the way for a more understanding and explanation of the study.

Many studies have been conducted to create a communal understanding of the elusive concept of entrepreneurship. It has been a subject of much debate and is defined differently by different researchers of different disciplines over time. To mention a few from the modern times of Schumpeterian, it is the act of innovation and implementation of change through carrying out of a new combination of resources (Bygrave & Zacharakis, 2011). It is an examination of how goods and services are discovered, by whom, and with what opportunities (Shane & Venkataraman, 2000). It is a dynamic process of vision, change, and creation (Kuratko & Hodgetts, 2007). Bosma (2013) defined entrepreneurship as any attempt at new business or new venture creation, such as self-employment, new business organization, or the expansion of an existing business, by an individual, a team of individuals, or an established business. Drucker (1985), Barot (2015), Chang and Wyszomirski (2015), and many more stated entrepreneurship as a process of creation of new businesses or creating new ways of managing the existing one through scanning and exploiting opportunities (Baumol & Schilling, 2008; Cieslik, 2017; Stevenson & Jarillo, 1990). From these, this study has conceptualized entrepreneurship as the process of creating new business and entering a formal market.

The economic and social contribution of entrepreneurship is the other studies’ dimension and reached the conclusion that it considerably contributes to sustainable economic development. This fact is massively supported by both theoretical and empirical studies. It contributes to the economic growth by channeling new innovation into the market (Audretsch, 2002; Valliere & Peterson, 2009). Relative to the existing firms, new entrants come up with new ideas and supply variety that can have more value to the market (Koster et al., 2012). Creating jobs, increasing GDP, reducing poverty, and enhancing whole society welfare is associated contribution (Burke, 2011; Ivanovi-Djuki et al., 2018), although, at the same time, economic growth has a significant influence on the development of entrepreneurship (Casares & Khan, 2016; Sabella et al., 2014). It also contributes to the enhancement of productivity through innovative use of the existing resources or by creating new ways of exiting modalities (Fritsch, 2008).

Though common interests and motives toward the development of entrepreneurship may be almost equally there in all economic levels of nations in the world, there is a great difference in terms of the number of entrepreneurial firms actually starting their operation and registering success. To show few facts on the ground for comparison from World Bank business entry-level database (2020), the new business entry rate per 1000 working-age population (15–64 age) in Afghanistan was 0.21, in Burkina Faso was 0.33, in Ethiopia was 0.51, in Iceland was 9.88, in Kosovo was 4.00, in Malta was 17.8, in Peru was 3.75, and in Hong Kong was 28.59. Studies show that such variations are likely to be associated with differences in environmental easiness to create entrepreneurs. For instance, a review of literature done by Gedeon (2010) on the field concluded that variations in entrepreneurship development levels across countries are determined by economic conditions and institutions situated to govern business activities. Qualities of governance explained in proxy factors such as politics, economic and labor market, entry and leaving regulations, education system, and infrastructure in a given nation are claimed to take a wider notion (Reynolds, 2007). A study conducted by Ha et. al. (2016) in Asia, based on the World Bank database, reveals the fact that governance indicators such as qualities of regulation, accountability, and control of corruption create differences in entrepreneurship development across countries. Friedmana (2014) did the similar study on the area and conclude the same by taking quality of governance as determinant factor in the context of 149 countries from all income economies.

Although these studies revealed the relationship between governance and entrepreneurship, some of them concluded their finding based on data only from one or a few countries or regions (Deb, 2013; Ha et al., 2016), while others focused only on a few of the governance indicators (Gil, 2011). Only the study conducted by Groşanu et. al. (2015) considered all the worldwide governance indicators as factors in the context of all income economies. Furthermore, some of the studies in the area were conducted a decade back and their conclusion is less likely helpful to know the current scenario. With these discrepancies, the full picture of the effect of governance indicators on entrepreneurship development in the current situation is less likely concluded. Therefore, this study intended to fill the mentioned gap by considering all income economies as the target based on data obtained from World Bank’s worldwide governance indicators (WGI), new business entry-level, and economic development difference (control variable) into consideration.

The finding of this study will contribute more to the existing body of knowledge. To mention specifically, from the policymakers and entrepreneurship program developers’ side, there is a tremendous increase in demand for new business entry to their economy. This demand pressurizes them to know facts on how governance affects their entrepreneurship. Therefore, this study can help them gain a better understanding of the existing phenomenon and, eventually, make a policy framework that paves future directions. A certain number of basic ideas of this study are expected to be shared by other researchers. In this regard, the study can partly help them access information regarding how governance affects entrepreneurship, particularly in the formal sector, and carry out more effective empirical studies using quantitative modeling. The study will help both actual and potential entrepreneurs enhance their intention to engage and expand in activities of entrepreneurship by showing multifaceted factors related to entrepreneurship governance. These factors may affect their firm negatively or positively. That is depending on how the actors in charge are proactive toward them. Thus, the study will benefit these entrepreneurs as an information source.

Background literature

Measure of entrepreneurship

The first question in entrepreneurship research is how to define entrepreneurship for the purpose of comparative benchmarking. At present, there are broadly two approaches. The first tries to look at entrepreneurship from the process aspect, while the second is from the contribution aspect. That means both the process and contributions of entrepreneurship can be used to measure the development of entrepreneurship in a given economic society. The process measure represents how entrepreneurship occurs and what activities are performed. The contribution measure represents the extent to which new firms enter the market, why, and what outcomes are yielded. In a logical sense, these two measures are highly correlated.

Self-employment and new firm formation are commonly used measures (Faggio & Silva, 2014) of entrepreneurship growth. This finding also has a theoretical view that vibrant entrepreneurship lies at the root of economic development through employment creation, productivity growth, and innovation (Van Praag & Versloot, 2007a, 2007b). Many empirical studies used self-employment as the proxy for entrepreneurship studies. Self-employed workers, according to ILO (2014), are those working on their own account across four sub-categories (employers, own-account workers, member of producer cooperatives, and contributing family workers). However, self-employment shows a different manner of economic partaking than entrepreneurship in the form of new business formation.

The new firm formation is also called new venture creation or entry. Data on new business entry, which is often collected by country sources, has become increasingly more accessible and provides accounts of new business entities. New business is frequently considered as an appropriate measure for entrepreneurship (Henrekson & Sanandaji, 2014). Though countries are collecting and organized the data for new firms’ formation at their own level, largely standardized and comparable data are provided by World Bank Group Entrepreneurship Snapshot (WBGES). This database provides comparative data on new firm entry at the country level. It is noteworthy that due to differences in definitions and legal treatments of different private organizational forms, the World Bank Group provides information on new Limited Liability Company (LLC) registrations only. New business formation, as a measure of entrepreneurship growth, is appropriately named entry density. New business entry density is the ratio between the number of limited liability companies newly registered per 1000 people to the population of working ages (15 to 64) (Klapper & Love, 2010; Munemo, 2012). This study used this measurement to measure entrepreneurship.

Governance indicators

The environment where entrepreneurs and enterprises exist determines the level of firm creation. Studies elsewhere try to reveal how contextual entrepreneurial environments create the differences. Although many factors and their proxies have long been studied, literature related to this study is limited to the association between governance and entrepreneurship. Governance, according to Kaufmann et. al. (2009) is the system by which countries ensure their organizations’ responsibilities. This means it is an overarching framework by which organizations are operated, guided, and made accountable. Effective governance leads to economic growth by enabling the business environment (Huynh & Jacho-Chavez, 2009).

Data regarding government effectiveness in their governance system and practice has been collected since 2006 from the views of a large number of enterprises, citizens, and experts in all nations by the World Bank. In addition, the Bank has also organized the aggregate data from a variety of survey institutions, non-governmental organizations, and international organizations. The governance indicators emphasized by the World Bank Project are voice and accountability, government effectiveness, regulatory quality, rule of law, political stability, and corruption control (Kaufmann et al., 2009). Governance in this study context is about how countries perceive indicators such as voice and accountability, government effectiveness, regulatory quality, rule of law, political stability, and corruption control and act accordingly.

Governance and entrepreneurship

There are previous studies regarding the relationship between governance and relationship with inconsistent results. Studies conducted by Çule and Fulton (2013) indicates that new business entry in new or existing market demands a moderate level of bureaucracy and proper regulatory quality as well as corruption control. This finding is also supported by the studies conducted by Nistotskaya and Cingolani (2015) by stating bureaucratic structure has an indirect effect on entrepreneurship rates through better regulatory quality. As suggested by Jalilian et. al. (2006) working on the improvement of regulatory quality improves business performance in particular and economy in general.

The relationship between business environment and entrepreneurship activity by Klapper et. al. (2007) condescending governance as the main pillar in the business environment and dedicated that higher percentage of firm registration and entry were experienced in countries with better governance. This argument was also supported by studies conducted by Amoros et. al. (2013), Dau and Cuervo-Cazurra (2014), and Dabija et. al. (2014). Their common argument was that fostering entrepreneurial activity can be stimulated by an effective regulatory framework, clearly defined property rights, transparent and easy procedure of business registration and entry. They also added political stability into consideration. The relative influence of governance proxy factors on entrepreneurship can be mediating by the existing economic development difference of countries. The governance framework of countries with high economies stimulates more of formal entrepreneurship entry than informal one (Thai & Turkina, 2014). This has an implication that the majority of entrepreneurs enter the market of low-income countries are informal categories. This idea is also shared by the study conducted by Dau and Cuervo-Cazurra (2014). Many more studies found out that new entrant entrepreneurs are relatively large where countries with less costly procedures in their business establishment. They also noted that good institutional arrangements positively influence entrepreneurship. With respect to the level of economic development, interference for entrepreneurship development because of better governance is not always related. The study conducted by Nyarku and Oduro (2017) in Ghana supports this fact by stating that bureaucracy, inconsistent policy climate, unsupportive customs, and regulations, constricted monetary and credit policies, corruption, and excessive tax practice, workforce, and labor regulations were found to negatively affect business entrants.

Political instability and weak control of corruption are also claimed to be critical governance dimensions in influencing business establishment. A study conducted by Abu et. al. (2015) in West Africa shows that rising corruption and political instability contribute to business under development by affecting government revenue, production, investment, and income distribution. With the same token, Alonso and Garcimartin (2013) identified that the extent to which countries control corruption determines enterprise establishment and innovativeness. On the contrary, studies basing low-income countries such as Goedhuys et. al. (2016) mentioned that corruption would rather have lubricated and accelerated enterprise establishment and innovativeness. A study conducted by DiRienzo and Das (2015) based on the global innovation index found that the extent of political stability and the quality of corruption control determine the extent to which entrepreneurs enter the market. John and Johnson (2015) conducted on the same topic in Nigeria found out political instability has a high negative influence on the number of new businesses coming to the market. The negative effect of corruption stated by Avnimelech et. al. (2020) in Kenya reveals that countries with a high level of corruption usually face a low level of productive entrepreneurship.

Government effectiveness and rule of law are other factors claimed to be determinate of entrepreneurship activities of countries. Countries with effective government and considerable rule of law registered high entry of new business and economic growth. A study conducted by Sasmaz and Sagdic (2020) in countries of transition economies found that only government effectiveness influences the business entry level. Zhou et. al. (2020) revealed that both movement effectiveness and rule of law contribute a lot in influencing the development of business through accelerating economic growth. They also consider the influence of voice and accountability in their study. The relationship between government effectiveness and entrepreneurship is negatively related (Friedman, 2011).

The aforementioned theoretical and practical literature and many several studies demonstrate different results for the relationship between governance and entrepreneurship. Some indicate all indicators have an influence with different magnitudes, whereas others show the significant influence of some of the factors. Few of them show negative relationships of some of the factors or no relationship at all. The difference may be because of the sources of data or their methodology difference or the combination of these and other reasons. Some of the studies used nationally registered formal business, while others take both formal and informal business from their country’s data set. Few studies have used data formally registered business data from the World Bank data set ware as others used the general level of entrepreneurship data provided by the Global Entrepreneurship Monitor (GEM) of both formal and informal entrepreneurship.

The analysis of the previous studies supports this study to formulate the following research questions and hypothesis on which the study empirically tested the relationship based on the data obtained from the World Bank entrepreneurship index represented by formally registered new business and worldwide governance indicators.

• Q: Which characteristics of governance indicators most strongly influence entrepreneurship?

• H₀: Voice and accountability, government effectiveness, regulatory quality, rule of low, corruption control, political stability, and gross net income have no statistically significant positive effect on entrepreneurship.

Research design and methods

Research design

This study used explanatory research design which is best fit to test the association between or among variables based on the underlying hypothesis (Sekaran & Bougie, 2010; Kumar, 2005; Singh, 2006). It is more likely to use quantitative data (Nargundkar, 2008) and helped the researchers to draw objective conclusion from the finding (Malhotra & Birks, 2000).

Sources of data

A total of 266 countries is revealed under the heading of new business entry density data set of the World Bank Group (2020) of which 132 countries have more than 3 to 5 year recent data (2014–2018). Six countries were excluded, because their last 2 year observations were not registered. As a result, 126 countries were taken as case countries. Based on their income level, 19 countries are from low income, 27 are lower-middle, 34 are upper-middle, and 46 are high-income countries. The 5 year data for worldwide governance indicators of the case countries were sourced from World Bank (2020). Though data regarding governance has been collected and organized since 2006 by the World Bank, its usability for this study was determined based on the available data of the dependent variable (entrepreneurship). This works for the control variable, Gross National Income (GNI)-per capita.

Variables description

To respond to the objectives of the study, variables pertaining to entrepreneurship and governance were taken with acknowledgment and used. Entrepreneurship was measured using the data sourced from the World Bank Group survey. This report measures entrepreneurial activity performance around the world. This database provides annual data from 2006 to 2018 and includes cross-country time-series data on the number of newly registered businesses around the world (World Bank, 2021a, 2021b). Formal entrepreneurship in this study context is any economic unit of the formal sector incorporated as a legal entity and registered in the country’s registry. This performance indicator has long been widely used in literature to study entrepreneurship determinant factors (Dau & Cuervo-Cazurra, 2014). Entrepreneurship is considered in this study as the dependent variable.

To measure governance, the study used worldwide governance indicators developed by World Bank (2020). The main objective of this report is to measure the quality of governance using six governance indicators. They are considered by this study as independent variables. The study assumed the difference in the level of economic development among the countries as the possible predictor that affect the relationship between governance indicators and entrepreneurship and took it as a control variable. The variable is represented by GNI (gross net income) per capita that can be the potential source of creating and operating business undertakings. These data are also sourced from World development indicators (World Bank, 2020).

The variables, their descriptions, measures and sources of data are depicted in Table 1.

Table 1 Variable descriptions

Method of analysis

Because of the objective to be addressed and the nature of the data, the study used inferential analysis including Pearson correlation and multiple linear regression models. To make the data fit for the mentioned models and aligned to the scale of measurement of the independent variables, the indexed data of dependent variable which varies from 0.01 to 28.54 and the control variable were transformed using natural logarithmic transformation. Correlation analysis was used to find out the degree of relationships among variables. Multiple linear regressions model (OLS) was applied to explain the effect of governance indicators and GNI on entrepreneurship. Assumption tests for regression were made before applying the regression. The analysis was done using SPSS version 25. The model for multiple linear regressions is specified as follows.

$$Y = \beta_{0} + \beta_{1} x_{1} + \beta_{2} x_{2} + \beta_{3} x_{3} + \beta_{4} x_{4} + \beta_{5} x_{5} + \beta_{6} x_{6} + \beta_{7} x_{7} + \varepsilon ,$$

where Y is the entrepreneurship (DV); x₁ is the voice and accountability (IV); x₂ is the political stability and absence of voice (IV); x₃ is the government effectiveness (IV); x₄ is the regulatory quality (IV); x₅ is the Rule of Law (IV); x₆ is the corruption control (IV); x₇ is the GNI-per capita (control); β_0, β₁ ⋯ β₇ are coefficients of determination; ε is the error term.

Results

This part discusses aspects of the study findings. The logical sequence of analysis and presentation here under are tests for normality, correlation analysis, tests for multicollinearity, test for heteroscedasticity, and regression analysis and results.

Normality test

In this part, we discussed whether the continuous data for dependent variable is normally distributed and fit for linear regression analysis. To assure this test–retest were made using Kolmogorov–Smirnov and Shapiro–Wilk test model. The two rows in Table 2 shows the results.

Table 2 Normality test

As indicated in the first row of Table 2p values of normality test of both Kolmogorov–Smirnov and Shapiro–Wilk are less than 0.05 and concluded that the data are departed from the normal distribution and are not fit for linear regression model. Thus, to make the data approach to normal distribution, natural logarithmic transformation was applied. As indicated in second row of the table, the probabilities (sig values) of both tests are greater than 0.05. Thus, the null hypothesis is accepted and concluded that the data are not different from normal.

Pearson correlation analysis

Pearson correlation analysis was done among eight variables with continuous data including the dependent variable to check the magnitude and direction of relationship among variables. Suggested approaches to translate the magnitude of correlation coefficients are several. While many researchers probably agree that a coefficient of < 0.1 is a negligible and > 0.9 a very strong relationship, values in between are arbitrary and arguable. The rule of thumb cutoff points of correlation coefficients indicated in the work of Hair et. al. (2010) shows that the size of correlation and their description in absolute value term are 0.91 to 1 is ‘very strong’, 0.71 to 0.9 is ‘strong’, 0.41 to 0.7 is ‘moderate’, 0.21 to 0.4 is ‘weak’, and 0.01 to 0.2 is ‘very weak’ Thus, this study used the mentioned rule of thumb as a base for interpreting the finding in Table 3.

Table 3 Pearson correlation analyses

The results in Table 3 shows that the dependent variable (LnEntrepreneurship entry density) is strongly correlated with regulatory quality (r = 0.806, p < 0.01), corruption control (r = 0.787, p < 0.001) and LnGNI per-capita (r = 0.780, P < 0.01). It has moderate correlation with government effectiveness (r = 0.693, p < 0.01), political stability (r = 0.690, p < 0.01), rule of low (r = 0.678, p < 0.01) and voice and accountability (r = 0.466, p < 0.01). The correlation results among independent variables and the control variable show that voice and accountability has moderately correlated with all predictor variables (0.506 ≤ r ≤ 0.659, p < 0.01). Political stability is strongly correlated with corruption control (r = 0.734, p < 0.01) and moderately correlated with the rest independent variables and the control one. Government effectiveness has very strong correlation with regulatory quality (r = 0.902, p < 0.01) and rule of low (r = 0.952, p < 0.01) and is strongly correlated with corruption control (r = 0.850, p < 0.01) and LnGNI per capita (r = 0.876, P < 0.01). Regulatory quality is strongly correlated with corruption control (r = 0.851, p < 0.01), rule of law (r = 0.878, p < 0.01), and LnGNI per capita (r = 0.867, p < 0.01). Corruption control is strongly correlated with rule of law (r = 0.866, p < 0.01) and LnGNI per capita (r = 0.794, p < 0.01). Rule of law has strong correlation with LnGNI per capita (r = 0.819, p < 0.01). This indicates that changes in each of the variable are linearly associated with shift in another variable.

The stronger the correlation the more difficult it is to change one variable without changing another variable. As a rule of thumb, if the absolute values of correlation among or between independent variables exceed 0.8, the problem of multicollinearity is suspected and this creates difficulty for the regression model to estimate the casual relationship between independent and dependent variables (Shrestha, 2020). This condition is revealed in the above correlation analysis, Table 3. Thus, further collinearity diagnosis was detected as depicted in Table 4 using VIF (Variance Inflation Factors) to drop highly correlated independent variables. VIF is used to measure how much the variance of the estimated regression coefficient is inflated if the independent variables are highly correlated.

Table 4 Multicollinearity test

Multicollinearity test using VI

Table 4 shows weather there are similarities among independent variables that could affect the result of regression outcome. The table is depicted as Coefficient A and coefficient B. The first indicates how all six independent variables and one control (LanGNI per capita) variable are correlated. Based on the coefficient output (Collinearity statics), obtained VIF value of government effectiveness (18.285) and rule of law (16.26) have Collinearity problems and were dropped from the model. The second (B) indicates how Collinearity among variables is avoided after two highly correlated variables were dropped from the model since the result fall between 0 and 10 (Belsley, 1991; Dhakal, 2018) and can be concluded that there is no multicollinearity symptoms.

Heteroscedasticity test

Under here difference in residual variation of the observed data was examined. According to the study conducted by Khaled et. al. (2019) if the significant value is greater than 0.05, then there is no problem of heteroscedasticity. The idea is that the residual value does not increase with increasing values of independent variable (Table 5).

Table 5 Heteroscedasticity test

The F ratio in the heteroscedasticity (Table 5) tests shows whether the data has heteroscedasticity problem or not. The results in the table shows the value obtained is F (5, 120) = 60.21, p (0.973) > 0.05. This leads to conclude that there is no heteroscedasticity problem and the data are fit for linear regression.

Regression results

Model fitness determination

The model summary Table 6 reports the strength of the relationship between the model (governance indicators) and the dependent variable (entrepreneurship). The table provides R, R², and adjusted R² and standard error of the estimate, which can be used to determine how well the regression model fit the data.

Table 6 Model summery table

As revealed in Table 6, Multiple Correlation coefficient (R) of 0.846 indicates a good level of relationship predication between independent and dependent variables. The “R²”-coefficient of determination is the proportion of variance in the dependent variable that can be explained by the predictors. The results of coefficient of determination indicated that the five predictors including the control variable (LnGNI per capita) which are kept in the model explained 71.5% of the variance in the dependent variable. In addition, 28.5% of the variation was caused by factors other than th predictors included in the model.

Statistical significance of the model

The F ratio in the ANOVA (Table 7) tests whether the overall regression model is a good fit the data. The results in the table shows that the predictor variables and the control variable statistically and significantly predict the dependent variable, F (5, 120) = 60.21, p (0.000) < 0.05. That means the regression model is a good fit of the data.

Table 7 ANOVA table

Statistical significance of the predictors

Statistical significance of each of the independent and control variables tests whether the unstandardized coefficients are equal to zero. That means for each of the coefficients, H0: β = 0 versus Ha: β ≠ 0 was conducted to investigate if each variable need to be in the model.

As indicated in Table 8, the t value and the corresponding p value are in their respective column. The tests indict us that political stability p (0.030) < 0.05; regulatory quality p (0.006) < 0.05; corruption control p (0.016) < 0.05, LnGNI per capita (control) p (0.037) are significant, but voice and accountability is insignificant p (0.305) > 0.05. This means that the independent variable, voice and accountability, is no more useful in the model, because it does not add a significant contribution in predicting the variations in the dependent variable (entrepreneurship).

Table 8 Coefficient table

Estimated model coefficients

The general form of equation to predict entrepreneurship development from political stability, regulatory quality, corruption control, and LnGNI per capita (control) is

$$\begin{aligned} {\text{Predicted}}\;{\text{Entrepreneurship}}\;{\text{development}} & = - 1.959 + 0.343\left( {{\text{political}}\;{\text{stability}}} \right) + 0.666\left( {{\text{regulatory}}\;{\text{quality}}} \right) \\ & \quad + 0.489\left( {{\text{corruption}}\;{\text{control}}} \right) + 0.192\left( {{\text{GNI}}} \right). \\ \end{aligned}$$

The constant − 1.959 is the predicted value for the dependent variable if all the predictors including the control are equal to zero: political stability = 0, regulatory quality = 0, corruption control = 0, and LnGNI per capita = 0. That is, we would expect an average entrepreneurship density of − 1.959 registered when all predictor variables take the value of zero.

Each unstandardized coefficient in the model indicates how much the dependent variable varies with an independent variable when all other independent variables are held constant. This means for every one unit increase in perception of political stability efficiency, there is 0.343 increases in entrepreneurship entry density. For every one-unit increase in perception of regulatory quality efficiency, there are 0.666 increases in entrepreneurship entry density. For every one-unit increase in perception of corruption control efficiency, there are 0.489 increases in entrepreneurship density. For every one-unit increase in GNI per capita performance, there are 0.192 increases in entrepreneurship entry density.

Conclusions and discussion

This study was intending to investigate the effect of governance indicators on entrepreneurship. Multiple linear regression mode was run to analyze the relationship. Before running the model, the pretest of the data suggested that the distribution of the dependent variable was highly skewed and data transformation was required and made. After the raw data were transformed using natural logarithms, assumption tests for normality, multicollinearity, and heteroscedasticity were made and the results assured the possibility for running multiple linear regression.

The coefficient of determination result of the regression analysis is 0.715. That means all independent variables including the control variable predict 71.5% of the variation in the outcome. The result of the overall regression model is F (5, 120) = 60.21, p (0.000) < 0.05. This result indicates that the model statistically and significantly predicted entrepreneurship. The result of each predictors are: political stability p (0.030) < 0.05; regulatory quality p (0.006) < 0.05; corruption control p (0.016) < 0.05, GNI (Gross Net Income) p (0.037) < 0.05; and Voice and accountability p (0.305) > 0.05.

In conclusion, the results of this research provide empirically supported models for the notion that some governance indicators are significantly influencing entrepreneurship. To make specific and show their magnitude of influence, out of the predictors, political stability, regulatory, corruption control, and GNI added statistically significant prediction capacity to the model with different degrees of contribution. The highest contributing predictor is the regulatory quality (0.666) and followed by corruption control (0.489). Political stability (0.343) and GNI (0.192) are contributing 3rd and 4th. Thus, the study suggests a number of implications for policymakers, program developers, and academics.

The ability of the government to formulate and implement sound entrepreneurship and regulations that permit and promote private sector development can create the difference in inviting and actualizing new entrant entrepreneurs into their market. When we say enabling policy and regulatory framework, they do mean ease of starting and closing the business and ease of registering the property. The strongest predicting effect of this variable is supported by the study conducted by Nyarku and Oduro (2017), and Dau and Cuervo-Cazurra (2014). Government subsidies that keep uncompetitive business alive create the difference. The issues of tax management and accessibility to finance are also very critical issues in motivating and joining entrepreneurs into the market.

Political stability and absence of violence have strong implication for the business entry, because it busts the perception and confidence of both investors, financial institutions, as well as the customers in the business chain. This implication is also supported by the study conducted by Shumetie and Watabaji (2019), and John and Johnson (2015) by stating political instability in a given nation is resulting in decreasing the extent of business innovativeness and new business entry in the market. In due course, it has its own big implication to entrepreneurship policymakers and program developers.

The difference in the extent of corruption control has a positive predicting effect on creating entrepreneurship. That means the more efficient the countries are in controlling corruption, the better in attracting new business entrants and maintaining the existing ones. This has its own implication for actors in the business world. The finding is supported by the studies conducted by Shumetie and Watabaji (2019), and Avnimelech et. al. (2020) by stating that in a society where corruption control is insignificant resulting in an increase in the cost of doing business and number of new entrants to the market. Countries suffering from a low and unsatisfactory level of entrepreneurship are known for low scores in corruption control experience. A study conducted in Guinea by Borowski (2017), for example, shows that the country was ranked 142 in 175 in corruption in 2016, its rank in entrepreneurship was 122 in 137 (Global entrepreneurship Index Bank, 2019). This also works for other similar countries such as Chad, Malawi, Mali, and Ethiopia from Africa. Firozjaii (2012) investigated that corruption has a regressive effect on entrepreneurship development, because it can make a country’s system benefit a few existing and better connected firms and create negative incentives for entrepreneurs. In countries where there is a loose entrepreneurial governance system, windows of entrepreneurial development opportunities are negligible for entrepreneurs.

The issues of voice and accountability have no substantial contribution in predicting entrepreneurship when the other predictors are in the model in this particular study. This result is in line with the proposed null hypothesis. This indicates that participation of citizen’s in the selection of the leaders, freedom of expression and association, and availability of free media are less likely to go with the density of new business entry in the market. This does not mean that it has no influence at all. It may have an indirect influence that needs to be taken into consideration while crafting or modifying their policy frame. In addition, while conducting such a study, variables beyond the study’s conceptual framework may affect the relationship between predictors, and the outcome variable. Considering this, this study has taken the difference in economic development level represented by gross net income as a control variable. The finding supports the argument. It is also supported by studies conducted by Stoica et. al. (2020). This has its own implication for further study and actors in the business chain. The higher the economic development of nations, the better in inviting business to their market because of the financial and infrastructural support they provide to their startups.

Limitations and future research indications

Some limitations are revealed by the fact that some empirical analysis could not be performed due to data unavailability for a large number of countries and the generalizability of the result did not include these countries. For instance, the study focused only on available formal entrepreneurship data collected and organized by World Bank as limited liability companies. The result would be better if the data for all countries are available and used for the study. Informal entrepreneurship was not incorporated. The study did not consider the success of the new business over time as a measure of the dependent variable. It would be quite important to investigate entrepreneurship development using the success as an outcome variable. Case countries are different in their legal origin. Because of the lack of data for the mentioned variable, the study did not use this difference as a controlling variable in the model. Thus, it is quite important to the test rule of law as a factor by considering their legal origin. To demonstrate how effective and accountable they are, few nations may provide fabricated data about the level of corruption prevalence and other associated sensitive topics that could affect the output of the study. Such incidence, which would be outside of the researcher’s control, was reduced in the study using World Bank data.

Pertaining to the predictors, the study was limited to governance indicators stipulated by World Bank. Further studies will be required to explore the influence of other macro-level factors on entrepreneurship. The relationship between variables was computed using multiple linear regressions. Other methods may be equally important or would bring better results. Thus, methodological change is recommendable to see the relationship further from a different perspective. In addition, how to make the governance indicators practical is also beyond the scope of this study. Thus, future research will also be required to investigate how these indicators are practically exercised.