Introduction

Entrepreneurship plays a vital role for production and employment generation (Luiz, 2010), and helps in overall economic growth and development (Ács et al., 2012; Ács & Naudé, 2013; Harper, 2003). Entrepreneurship is also stimulated by innovation (Audretsch et al., 2006), and as a result, creative ideas thrive through entrepreneurship. Considering its importance, United Nations sustainable development goals (SDGs) have set targets (SDGs 8.3 and 9.3) for entrepreneurship promotion to achieve sustainable economic growth and inclusive industrialization for the growth of small and medium start-ups(United Nations, 2022).

Most developed countries have a good and supportive entrepreneurial ecosystem for start-ups (Ács et al., 2019). However, being the world’s third largest economy (World Bank, 2022) Japan does not have a satisfactory level of entrepreneurship (Ács et al., 2019; Bosma et al., 2020; IMD, 2022). According to the Global Entrepreneurship Index (GEI) in 2019, Japan ranked 26th out of 137 countries with unstable pillars of entrepreneurship (Ács et al., 2019). Furthermore, according to the Global Entrepreneurship Monitor (GEM) report 2019, entrepreneurial attitudes are also at the lowest level in Japan among the developed countries (Bosma et al., 2020). At the same time, Japan is ranked 29th in the World Bank ease of doing business index 2020 (World Bank, 2020). Similarly, Japan ranks 34th out of 63 countries in the IMD World Competitiveness Yearbook (IMD, 2022). Despite its economic strength, Japan ranks low in all international reports on entrepreneurship compared with its peer economies (USA, UK, China, Germany, and France). However, entrepreneurship was very crucial for Japanese development during 1960-1970s (Watanabe, 1999), and throughout history, the Japanese economy has prospered greatly due to the contributions and endeavors of its entrepreneurs (Aoyama, 2009). Moreover, since the burst of the economic bubble in the 1990s, Japan has experienced a decline in the rate of new business startups (Honjo, 2015). As a result, it is necessary to identify and explain the gap in recent entrepreneurial activities in Japan compared to the countries with similar economic strength. Hence, we hypothesized that the entrepreneurial activities in Japan are same as the entrepreneurial activities in peer economies, and we tested this hypothesis by using Oaxaca-Blinder decomposition analysis.

Honjo (2015) was the first to examine the factors that affect the entrepreneurial activities in Japan (Naiki & Ogane, 2021). However, Honjo (2015) did not examine the factors behind the entrepreneurial gap in Japan compared to peer countries. Consequently, our study aimed to bridge this gap in literature by identifying the factors behind the entrepreneurial gap in Japan compared to its peer countries. In our study, the Adult Population Survey (APS) data for the period of 2001–2018 from the Global Entrepreneurship Monitor (GEM) was used to identify the major entrepreneurial attitudes behind the low level of entrepreneurship in Japan compared to five other developed countries: the USA, UK, China, Germany, and France. We have utilized data from GEM due to its distinction as the most extensive and exclusive data source providing individual entrepreneur level information on a country-by-country basis.

The Oaxaca-Blinder decomposition developed for non-linear model (Yun, 2004) was used to identify the attitudinal factors for entrepreneurial activity gap in Japan compared to other similar economies. Because, when dealing with binary dependent variables, employing ordinary least squares (OLS) for Oaxaca-Blinder decomposition leads to unreliable decomposition outcomes. As our three dependent variables are binary, so to address this issue, we applied Oaxaca-Blinder decompositions specifically designed for models with binary dependent variables by Yun (2004). We used three outcome variables to measure the entrepreneurial activities, namely Entrepreneurship, Investment, and Future Intention for Entrepreneurship. Usually three dimensions (attitudes, ability, and aspirations) are used to understand the entrepreneurial activities (Ács et al., 2019). However, the GEM data contains only country-level individual data on attitudes toward entrepreneurship. Hence, in our study, the entrepreneurial activity gap has been decomposed in terms of four entrepreneurial attitudes (Network, Opportunity, Confidence, and Fear of Failure). In addition, we included a set of socio-demographic variables to control the economic and demographic variation of the respondents for further robustness check of our estimation.

The results show that three entrepreneurial indicators are lower in Japan compared to the majority of the peer countries studied, which implies that Japan has lower levels of Entrepreneurship, Investment, and Future Intention for Entrepreneurship, leading to the rejection of the null hypothesis of similar levels of entrepreneurial activities. The decomposition results show that Confidence shares most of the part (50–60%) of the explained portion of the Entrepreneurship gap in Japan compared to other countries. We can interpret that if Japanese people had same level of confidence like other countries, the Entrepreneurship gap would be minimized by 50–60%. Similarly, for the Investment gap in Japan, Confidence and Network explain most of the gap (about 60–70%), whereas Confidence and Opportunity are the mostly responsible (85–90%) for the gap in future business intention in Japan.

Our study contributes in at least three ways. First, to the best of our knowledge, this is the first study in literature which focused on the entrepreneurial attitudes behind the gap in entrepreneurial activities in Japan compared with other peer economies. Honjo (2015) was the first to examine the factors that affect the entrepreneurial activities in Japan. However, Honjo (2015) did not investigate the entrepreneurial activities gap and also the factors behind the entrepreneurial activities gap. Our study, on the other hand, measured the entrepreneurial activities gap in Japan compared with other countries as well as determined the factors behind the entrepreneurial gap in Japan. Second, our study is the first attempt to employ Oaxaca-Blinder Decomposition method to identify the entrepreneurial attitudes behind the entrepreneurial activities gap in entrepreneurship literature. In contrast, Honjo (2015) used logistic regression model to identify the factors associated with the entrepreneurial activities in Japan. Third, our study used more recent and longer period data (2001–2018) which can give additional insights about the entrepreneurial activities in Japan. In contrast, Honjo (2015) used same dataset for the period of 2001–2012 in his study.

This research offers valuable insights for policymakers aiming to promote entrepreneurship in Japan and other comparable economies. Implementing the policy implications can foster sustainable economic growth and help in accomplishing the objectives of sustainable development goals.

The sections are organized as follows: Section 2 describes data and methods; Section 3 presents results and discussion. Section 4 concludes the paper.

Data and method

Data

This study utilized data from the renowned Global Entrepreneurship Monitor (GEM) Adult Population Survey (APS) spanning from 2001 to 2018. Since 1999, GEM has been a joint project of Babson College (USA) and London Business School (UK) for entrepreneurship research. The GEM project seeks to identify the societal values, individual attitudes, and entrepreneurial framework that underpin the development of entrepreneurship (Bosma et al., 2021). Adult population survey is a regular survey of GEM research project to study the aspects of entrepreneurial behavior like characteristics and attitudes for new start-ups. Every year, GEM collects around 2000 random samples from various segments of adult people (aged 18 to 64) in each country and region. The cross-section data for each country from 2001–2018 were combined and made a pooled cross-section data for our analysis.

Outcome variables

This study fitted models for three entrepreneurial indicators —Entrepreneurship, Investment and Future Intention for Entrepreneurship– to identify the entrepreneurial attitudes behind the gap in entrepreneurial activities in Japan. In GEM data, Entrepreneurship has been measured by the variable total early-stage entrepreneurship which investigated the question of involvement with a nascent or young firm. In addition, the assessment of Investment and Future Intention for Entrepreneurship involved inquiries about the funding allocated for a new business and the anticipation of commencing a fresh entrepreneurial venture, respectively. Details of the outcome variables are shown in Table 1.

Table 1 Description of Outcome Variables
Full size table

We have compared entrepreneurial activities in Japan with five other peer countries which have progressed in entrepreneurial activities, namely USA, UK, China, Germany, and France. The distribution of sample size for each country is presented in Table 2.

Table 2 Distribution of Sample Size in each Country over the Period
Full size table

Figure 1 indicates that entrepreneurial activities i.e., Entrepreneurship, Investment, and Future Intention for Entrepreneurship in Japan are on average lower than other compared countries.

Fig. 1
figure 1

(Source: Author’s own calculation from GEM APS data 2001–2018)

Average Entrepreneurship, Investment, and Future Intention in Study Countries

Full size image

Explanatory variables

This study has considered four variables as key explanatory variables about perception and attitudes toward entrepreneurship: Network, Opportunity, Confidence (about knowledge, skill, and experience), and Fear of failure. In this case, Network has been measured by the question of knowing someone who has started a business, and the Opportunity was calculated by asking about prospect of starting new business. At the same time, Confidence and Fear of failure have been assessed by asking personal opinion about their self-belief and fear. Network is an important element of entrepreneurship, as network enhances the business performance (Chell & Baines, 2000, Spigel et al., 2018). Opportunity also promotes entrepreneur based economic development (Dhaliwal, 2016). Confident people have more tendencies to start ventures (Hayward et al., 2006). On the other hand, fear of failure hinders the entrepreneurship as a negative motivation (Cacciotti et al., 2016). So, we considered these four important entrepreneurial attitudes as our explanatory variables to describe the entrepreneurial gap in Japan.

For robustness check, along with these four variables, socio-demographic variables (age, gender, education years, and household income) have been also considered as explanatory variables to control individual characteristics. Household income variable has three categories: lower 33%, middle 33% and upper 33% of the country. This study has used dummy for the household income and the lower 33% dummy was considered as the base category for comparison. Table 3 displays the description of the explanatory variables used in this study.

Table 3 Description of Explanatory Variables
Full size table

Table 4 shows the descriptive statistics of the key outcome variables and explanatory variables for every country for the years 2001 to 2018.

Table 4 Descriptive Statistics of the Key Explanatory Variables
Full size table

As shown in Table 4, Japan has low average for entrepreneurial activities than other countries in this list. China has the highest average of the variable Entrepreneurship- 14% of the respondents are involved in entrepreneurship in China. On the other hand, Japan has the lowest average of Entrepreneurship—only 3.4% of respondents are involved in entrepreneurship in Japan. Also, in two other entrepreneurial activities Japan has poor average, in terms of Investment the average is 0.012 and in terms of Future Intention of Entrepreneurship, the average is 0.049. Considering this low level of average of entrepreneurial activities, this study attempted to find the gap of entrepreneurship in Japan in terms of the previously mentioned four entrepreneurial attitudes (Network, Opportunity, Confidence, and Fear of Failure). Moreover, this study has decomposed the gap to identify which attitude is more responsible for the entrepreneurial gap in Japan compared to other countries.

Method

This study aims to identify the entrepreneurial attitudes that are more influential behind the gap in entrepreneurial activities in Japan compared to other peer countries. Oaxaca-Blinder decomposition method (Blinder, 1973; Oaxaca, 1973) is the most widely used technique for examining the factors behind the differences in predicted mean between two groups by decomposing the mean difference into explained and unexplained differences– one portion is explained by variations in observed characteristics and another portion ascribable to differences in the estimated coefficients. In Oaxaca-Blinder (O-B) decomposition, the dependent variable is continuous and decomposition techniques have primarily been utilized within the framework of linear regression models.

However, in many cases, the dependent variable takes on a binary form. In such cases, employing the Oaxaca-Blinder decomposition with ordinary least squares (OLS) leads to inconsistent estimations of parameters, consequently produces misleading decomposition outcomes. Numerous research works have developed and utilized Oaxaca-Blinder decomposition techniques for models involving binary dependent variables. (Even & Macpherson, 1990; Fairlie, 1999, 2005; Gomulka & Stern, 1990; Yun, 2004).

Yun (2004) developed an expansion of the Oaxaca-Blinder decomposition technique for non-linear regression models involving binary dependent variables. Since, our three dependent variables are binary variable; we used Yun’s (2004) approach in our study, which is explained below:

Let's consider a binary dependent variable \(\mathrm{Y}\), which is determined by a linear combination of independent variables \(\mathrm{X}.\) The relationship is expressed as follows:

$$Y=F\left(X\beta \right)$$

Here, \(Y\) is a \(Nx1\) vector representing the binary dependant variable, \(X\) is a \(NxK\) matrix of independent variables and \(\beta\) is a \(Kx1\) vector of coefficients. The function \(F\) maps the linear combination of \(X\left(X\beta \right)\) to \(Y\), and the function \(F\) is required to be differentiable once. In this context, we can decompose the difference in the predicted mean of \(Y\) between groups \(i\) and \(jp\) as follows:

$$\overline{{Y }_{i}}-{\overline{Y} }_{{jp}}=\left[F\left({\overline{X} }_{i}{\beta }_{i}\right)-F\left({\overline{X} }_{jp}{\beta }_{i}\right)\right]+ \left[F\left({\overline{X} }_{jp}{\beta }_{i}\right)-F\left({\overline{X} }_{jp}{\beta }_{jp}\right)\right]$$
(1)

In the equation, \({Y}_{i}\) represents the predicted mean of \(Y\) for group \(i\) and \({Y}_{jp}\) represents the predicted mean of \(Y\) for group \(jp.\) The decomposition analysis is conducted for each combination of \(i\in \{USA,UK,China,Germany,France\}\) and Japan (\(jp\)). In other words, the model is estimated separately for each pairing of \(\{i,jp\}\).

The aforementioned decomposition is made at the aggregate level to decompose the variations in the predicted mean by considering the differences in characteristics \(\left[F\left({\overline{X} }_{i}{\beta }_{i}\right)-F\left({\overline{X} }_{jp}{\beta }_{i}\right)\right]\) and by considering the differences in coefficients \(\left[F\left({\overline{X} }_{jp}{\beta }_{i}\right)-F\left({\overline{X} }_{jp}{\beta }_{jp}\right)\right]\).

To determine the proportionate contribution of individual variable on the predicted mean gap, considering both attitudes and coefficient effects, we applied a decomposition formula introduced by Yun (2004). Following Yun (2004), a detailed decomposition equation of Eq. (1) can be expressed as follows:

$$\overline{{Y }_{i}}-{\overline{Y} }_{{j}_{P}}=\sum \nolimits_{i=1}^{K}{W}_{\Delta X}^{i}\left[F\left({\overline{X} }_{i}{\beta }_{i}\right)-F\left({\overline{X} }_{jp}{\beta }_{i}\right)\right]+\sum \nolimits_{i=1}^{K}{W}_{\Delta \beta }^{i}\left[F\left({\overline{X} }_{jp}{\beta }_{i}\right)-F\left({\overline{X} }_{jp}{\beta }_{jp}\right)\right]$$
(2)

where,

$${{W}^{i}}_{\Delta X}={\frac{\left({\overline{X} }_{i}^{i}-{\overline{X} }_{jp}^{i}\right){\beta }_{i}^{i}}{\left({\overline{X} }_{i}^{i}-{\overline{X} }_{jp}^{i}\right){\beta }_{i}^{i}}}, {{W}_{\Delta \beta }^{i}}=\frac{{\overline{X} }_{jp}^{i}\left({\beta }_{i}^{i}-{\beta }_{jp}^{i}\right)}{{\overline{X} }_{jp}^{i}\left({\beta }_{i}-{\beta }_{jp}\right)}\;\mathrm{and}\sum \nolimits_{i=1}^{K}{W}_{\Delta X}^{i}=\sum \nolimits_{i=1}^{K}{W}_{\Delta \beta }^{i}=1$$

As our dependent variables are binary choice variable, and \(Pr\left(Y=1\right)=F\left(X\beta \right)\), where \(F\) is a logistic cumulative distribution function.

Once the coefficient estimates are available, determining the weight becomes an easy task by utilizing the mean values of characteristics along with their respective coefficients.

Results and discussion

To identify the factors behind the entrepreneurial activities (Entrepreneurship, Investment, and Future Intention for Entrepreneurship) gap in Japan compared with other similar economies, we have estimated the results in two steps. First, we estimated the Logistic Regression model for the three outcome variables, and then estimate O-B decomposition results Eq. (1) and Eq. (2).

Determinants of entrepreneurial activities

In this study, three logistic regression models have been fitted for three entrepreneurial activity indicators: Entrepreneurship (Table 5), Investment (Table 6), and Future intention to become an entrepreneur (Table 7) to describe the association between dependent variables and explanatory variables. The logistic regression technique was employed to identify the factors influencing entrepreneurial activities, encompassing Entrepreneurship, Investment, and Future Intention for Entrepreneurship.

Table 5 Logistic Regression Results for Entrepreneurship
Full size table
Table 6 Logistic Regression Results for Investment
Full size table
Table 7 Logistic Regression Results for Future Intention for Entrepreneurship
Full size table

The results from Tables 5, 6 and 7 conclude that Fear has negative association with all three dependent variables – the more Fear, the less entrepreneurial activities which coincides with the findings of (Cacciotti et al., 2016). Again, Network has positive and significant relationship on Entrepreneurship, Investment, and Future intention of Entrepreneurship; it can be inferred that people having strong networks might be more involved in entrepreneurial activities. This finding is supported by the previous study (Chell & Baines, 2000; Spigel et al., 2018). In addition, Opportunity also has positive and significant association with all three entrepreneurial indicators, but the impact is lower compared to Network and Confidence. This result coincides with (Dhaliwal, 2016). Confidence has the highest positive as well as significant relation with every entrepreneurial activity which is also supported by (Hayward et al., 2006).

Tables 5, 6 and 7 show that, among the socio-demographic indicators, age is positively associated with Entrepreneurship and Future Intention of Entrepreneurship which implies that Entrepreneurship as well as future intention for start-up increase with age. However, square of age (agesq) is negatively associated with both of them, which indicates that after a certain level Entrepreneurship and future intention for entrepreneurship tendency reduce with age. So, age has a reversed U-shaped relationship with Entrepreneurship and Future Intention of Entrepreneurship. This result (Entrepreneurship) complemented the findings of Honjo (2015). On the other hand, Investment decreases with the increase of age but after a certain age Investment rises with the increases of age indicating a U-shaped relationship with age and Investment.

In Japan, females invest in a new business started by someone else more than male (Table 6). In contrast, females are lagging behind in all other entrepreneurial activities than males in all countries in this study. The relationship between entrepreneurial activities and years of schooling and household income varies across the countries and entrepreneurial activities.

In the next section, we have presented the Oaxaca-Blinder decomposition result to identity the significant entrepreneurial attitudes behind the entrepreneurial activities gap in Japan compared with other similar economies.

Decomposition of gaps in entrepreneurial activities

Decomposition of gaps in entrepreneurship

Oaxaca-Blinder (O-B) Decomposition of Entrepreneurship gap between Japan and compared countries are presented in Columns (3)-(7) in Table 8. Column (2) represent the predicted mean of Entrepreneurship in Japan.

Table 8 Oaxaca-Blinder Decomposition Analysis for Entrepreneurship
Full size table

In Panel A of Table 8, the predicted mean of Entrepreneurship represents the percentage of people engaged in entrepreneurial activities in each country. Among the countries, the predicted mean of Entrepreneurship in Japan is around 4.5%, the highest mean Entrepreneurship is in USA (12.6%) and the lowest mean Entrepreneurship is in France (4.13%). The Predicted mean of Entrepreneurship for all countries are highly significant at 1% level of significance. Oaxaca-Blinder decomposition technique finds the gap between two countries, presented in the third row of Table 8 (Ent. Gap in Japan) indicates the mean difference of Entrepreneurship between Japan and other countries. The mean difference of Entrepreneurship between USA and Japan is around 8% (Predicted mean of USA—Predicted mean of Japan = 0.0802). This difference suggests that the average Entrepreneurship in USA is 8% higher compared to Japan. The mean difference of Entrepreneurship gap in Japan compared with UK, China, and Germany is positive and statistically significant whereas the mean difference of Entrepreneurship gap between Japan and France is negative but insignificant. The mean differences of Entrepreneurship gaps indicate that all the compared countries except France have higher level of Entrepreneurship than Japan.

Using O-B decomposition technique, the total mean difference can be divided into two parts: explained and unexplained difference. Explained difference is due to the difference in explanatory variables. In Panel B of Table 8, explained difference accounts for the major portion of the gap in Japan with other countries and all explained differences are highly significant. The unexplained difference is due to the differences between coefficients, or it can be said the portion of difference that is not explained by the explanatory variables. All the unexplained differences in Table 8 are very small.

Following (Yun, 2004), we have identified the contribution of each explanatory variables on the explained gap of Entrepreneurship in Japan. Our estimated results show that among the explanatory variables, the Confidence (having knowledge, skill and experience) shares the major portion (about 60%) of explained difference of Entrepreneurship gap in Japan compared to other peer countries, except China (about 32%). For example, Confidence shares 60% of the total explained differences of Entrepreneurship gap between UK and Japan (Column 4). As the explanation of Oaxaca-Blinder decomposition, it can be inferred that if the Japanese people had confidence like the UK people the Entrepreneurship gap in Japan would be minimized by 60%. Our findings coincide with the results of Honjo (2015) and (Naiki & Ogane, 2021). Honjo (2015) found that Japanese people having knowledge, skill and experiences have higher tendency to start business than the people of other countries. Naiki and Ogane (2021)also found that the Japanese full-time worker having more confidence are less likely to initiate a business venture in Japan compared with non-full-time worker.

Similarly, we found that Confidence contributes more than 55% of the explained part of Entrepreneurship gap in Japan with the USA, and Germany. The results indicate that Japanese people feel they have less confidence for start-ups. However, Network is found to be more responsible for explaining the Entrepreneurship gap between China and Japan – which is about 47% followed by the share of Confidence (32%).

In summary, the results show that Confidence is the most important factor for minimizing the entrepreneurship gap in Japan compared to other countries.

Decomposition of gaps in investment

Table 9 presents the predicted mean gap of Investment between Japan and other countries as well as the decomposition results of the gap partitioned into explained and unexplained differences. Column (2) in Panel A of Table 9 shows that, on average 1.7% of the people have invested in business started by someone else other than stock market in Japan indicating that around 1.7% of the respondents are investor in Japan. Column (3)-(7) in Panel A of Table 9 indicate that among the compared countries, the percentage of investor is highest in China (about 8.8%), followed by USA (6.3%), Germany (5.1%) and lowest in Japan (1.7%). The predicted mean Investment gap in Japan compared to other countries are presented in Column (3)-(7) in Panel A of Table 9. The predicted mean Investment gap is the difference of Investment between other countries and Japan. We found that this mean difference is positive for all the comparing countries which suggests that the level of Investment in Japan is lower than all the countries in this analysis. On average, China has the highest level of Investment compared with Japan, which is 7% higher whereas the mean difference of Investment between UK and Japan is the lowest (i.e., 0.7% higher investment in UK than Japan). All mean differences are statistically significant.

Table 9 Oaxaca-Blinder Decomposition Analysis for Investment
Full size table

Panel B of Table 9 presents the decomposition results of the total mean difference between other countries and Japan into explained and unexplained differences. The results show that all the explained differences between other countries and Japan are statistically significant. Among the explanatory variables, Confidence and Network are mostly responsible for the disparity in Investment in Japan compared to other countries. For instance, Confidence explains around 47% and Network explains around 27% of the total explained gap of Investment between Germany and Japan. Similar results are found for the USA, UK, and France, i.e., Confidence and Network are accountable for explaining most of the part of explained gap of Investment between other countries and Japan. It can be interpreted that if Japanese people had Confidence and Network like other countries, the large portion of the gap for Investment would be minimized. This table leads us to the conclusion that Confidence and Network have the highest magnitude for the disparity of Investment in Japan than other countries. Our result is supported by Honjo (2015), where using the similar data set for the period of 2001–12 he concluded that people having connection with other entrepreneurs tend to invest more in Japan than other countries. However, with additional six years data (2013–2018) after Honjo’s (2015) study, our study found that both Confidence and Network are vital for minimizing the Investment gap in Japan.

On the other hand, the unexplained gap of differences for Investment for most of the countries (except for the UK) are significant and account for a considerable amount.

In short, the above results indicate that both Confidence and Network are the most important factors for minimizing the investment gap in Japan compared to other peer countries.

Decomposition of gaps in future intention of entrepreneurship

Though we cannot directly measure the entrepreneurial activity through future intention of respondents about starting new business in near future, we can only shed light on the tendency of the plan to establish a new startup within the upcoming three years. In Panel A, Column (2) of Table 10 shows that the predicted mean of future intention of starting a new business is 7% in Japan. From Column (3)-(7) of Table 10, we found that China has the highest level of intention (21%), whereas UK has the lowest level of intention (7%) to establish a new startup within the upcoming three years. Mean gaps of Future intention of Entrepreneurship between other countries and Japan are positive and statistically significant for all countries except for UK. It implies that people from other countries (except UK) have more intention to start new enterprises than Japan. As a result, all countries except UK in this Table 10 have higher level of intention for start-ups than Japan.

Table 10 Oaxaca-Blinder Decomposition Analysis for Future Intention for Entrepreneurship
Full size table

The Panel B of Table 10 shows the results of Oaxaca-Blinder decomposition for the plan of starting a new startup in the next three years. We found that Confidence along with Opportunity is the important factor for future business plan. For example, Confidence is responsible for 55% (followed by the Opportunity 37%) of the gap between France and Japan. The interpretation is, if Japanese people had Confidence and Opportunity like France, this gap of future intention in Japan would be minimized mostly. Confidence is responsible for the gap of Future intention of Entrepreneurship in Japan compared to USA, UK, Germany, and France which is around 50%. Similar to the other two previous entrepreneurial activities, Network in China plays the most important role in explaining the explained gap of mean differences (about 53%) than other explanatory variables. Confidence and Opportunity are the vital factors for explaining the mean gap of future intention of business between other countries and Japan in all compared countries except China.

However, the unexplained gap of mean differences for future intention of starting a new business in next three years for most of the countries (except for the USA) are significant and account for a considerable amount.

In summary, the results suggest that both Confidence and Opportunity are the most important altitudinal factors for minimizing the gap of future intention of business in Japan compared to other similar countries.

Thus, the null hypothesis of similar level of entrepreneurial activities in Japan and other similar countries has been rejected for all of the three models with entrepreneurial attitudes (Network, Opportunity, Confidence, and Fear of failure).

Robustness check

To intensify the results of this study, we have checked the robustness of our results by including the socio-demographic variables in our models. The results for robustness checks are presented in Table 11, 12 and 13. The attitudinal variables (Network, Opportunity, Confidence, and Fear of failure) of entrepreneurship explain major portion of the explained gap for all three entrepreneurial activities, i.e., Entrepreneurship, Investment, and Future Intention for Entrepreneurship. The results of all the attitudinal variables (Network, Opportunity, Confidence, and Fear of failure) of entrepreneurial activities (Entrepreneurship, Investment, and Future Intention for Entrepreneurship) are consistent and robust with the main findings and statistically significant at 1% level.

Table 11 Oaxaca-Blinder Decomposition Analysis for Entrepreneurship using Socio-demographic Variables
Full size table
Table 12 Oaxaca-Blinder Decomposition Analysis for Investment using Socio-demographic Variables
Full size table
Table 13 Oaxaca-Blinder Decomposition Analysis for Future Intention for Entrepreneurship Using Socio-demographic Variables
Full size table

The socio-demographic variables (age, female, education years, and household income categories) have tiny share in explaining the explained gap for all three entrepreneurial activities.

Conclusion

Japan’s entrepreneurship environment is inadequate in comparison to its economic strength (Ács et al., 2019). Several reports and studies have concluded that Japanese entrepreneurship is not in a satisfactory position (Ács et al., 2019; Bosma et al., 2021; IMD, 2022). The Japanese economy thrived in the 1960s because of some exceptional entrepreneurship. Since the collapse of the bubble economy in the 1990s, the rate of new business entry has decreased in Japan Honjo (2015). However, to the best of our knowledge, no studies explored the factors behind the mean differences in the entrepreneurial activities in Japan compared with other similar economies. To address this gap, our study aimed to identify the entrepreneurial attitudes behind the mean differences in Japanese entrepreneurial activities (Entrepreneurship, Investment, and Future Intention for Entrepreneurship) compared with other similar economies.

This study analyzed the data on entrepreneurship collected from GEM on five countries apart from Japan (USA, UK, China, Germany, and France) for the period of 2001–2018. GEM is a joint project of Babson College (USA) and London Business School (UK) and is the largest and authentic data source for entrepreneurship research. We have analyzed the data in two steps. First, logistic regression model was used to identify the determinants of the entrepreneurial activities (Entrepreneurship, Investment, and Future Intention for Entrepreneurship). Second, the extension of Oaxaca-Blinder decomposition technique for non linear dependent variables (Yun, 2004) was used to determine the entrepreneurial attitudes (Network, Confidence, Opportunity, and Fear) behind the mean differences in Japanese entrepreneurial activities. To determine most influential entrepreneurial attitudes for the mean differences of entrepreneurial activities in Japan, we first used entrepreneurial attitudes (Network, Confidence, Opportunity, and Fear) as the explanatory variables. Then for the robustness check, we included socio-demographic variables in the models along with entrepreneurial attitudes.

Descriptive statistics of the outcome variables indicate that entrepreneurial activities (Entrepreneurship, Investment, and Future Intention for Entrepreneurship) are on average lower in Japan than other compared countries. In the first part of our analysis, the logistic regression results show that the more Fear, the less entrepreneurial activities. Again, people having strong networks might be more involved in entrepreneurial activities. Having good opportunity for start-up and Confidence (having knowledge, skill, and experiences) promotes entrepreneurial activities. From the second part of our analysis applying Oaxaca-Blinder decomposition, we found that the predicted mean differences of entrepreneurial activities between Japan and other countries confirm the existence of significant gap entrepreneurial activities in Japan. To identify the factors behind this gap, we have employed the Oaxaca- Blinder decomposition which provides us the explained and unexplained parts of total mean differences as well as detailed decomposition (share of each explanatory variables).

Findings from Oaxaca-Blinder decomposition can be summarized in three points. First, for the gap in Entrepreneurship in Japan compared to other countries, this study found that Confidence has the highest magnitude, which is responsible for 50–60% of the gap in Entrepreneurship in Japan with the comparing countries. According to Oaxaca-Blinder decomposition interpretation, it might be concluded that if Japanese people had same level of confidence like other countries, then the gap would be minimized by 50–60%. However, the result is different for the gap between Japan and China, in this case Network is more important than Confidence. Second, for explaining the gap in Investment in Japan, confidence along with network has the highest magnitude (60–70%). As a result, in this part it might be interpreted that if Japanese people had more Confidence and Network like other countries Japanese Investment gap would be minimized largely. Third, Future Intention of Entrepreneurship in Japan also has disparity with other countries, and for explaining this gap, Confidence and Opportunity showed highest magnitude (85–90%) among all explanatory variables. Similarly, this study might be interpreted that if Japanese people have more Confidence and Opportunity more entrepreneurs may appear in Japan.

The most important implication is that this study is the first attempt to identify the significant entrepreneurial attitudes for the entrepreneurial gap in Japan compared with other countries. This study provides important policy implications to enhance entrepreneurial activities in Japan as well as in other countries for sustainable economic development. Also, our paper extends the work of Honjo (2015) by incorporating more time periods and recent data as well as decomposition analysis and thus contributes to entrepreneurship literature.

This study has discovered a link between entrepreneurial attitudes and entrepreneurial activities. As a result, this research outcome can be used for developing policy tool for entrepreneurship improvement in Japan as well as in other countries. This research result can be used for future research on entrepreneurship gap minimization.

One of the limitations of this study is that this study has only analyzed the data on attitude towards entrepreneurship from the GEM data source. As a result, all three features of entrepreneurship (attitudes, ability, and aspirations) were not possible to consider for measuring the entrepreneurial activities. Moreover, no intervention for entrepreneurship development was possible to study the causal relationship of entrepreneurship and entrepreneurial attitudes. This study also could not consider the macroeconomic profile of the country such as GDP size, per capita income, unemployment rate, taxation and other variables due to lack of entrepreneurial ecosystem data which are important for entrepreneurship. Further research can be carried out considering all three aspects (attitudes, ability, and aspirations) of entrepreneurship to get accurate outcome. More research can be conducted with the intervention of entrepreneurship to determine the causal relationship between entrepreneurship and entrepreneurial attitudes. In addition, more research is suggested to consider country’s economic profile to get more precise result about entrepreneurial environment.