Exploratory Research on Relationship between Entrepreneurial Orientation Dimensions and Business Performance and Growth of Fast and Slow Growing Small and Medium Enterprises in Bosnia and Herzegovina

This paper analyses environment in fast (gazelles) and slow growing companies (mices) using Entrepreneurial orientation (EO) dimensions and business performance in terms of sales and employee growth. The research is done using information on 178 small and medium enterprises (SMEs) in Bosnia and Herzegovina. We find that EO dimensions are more present in gazelles than in mices. We show a small to moderate significant correlation between EO dimensions and business performance of gazelles and mices. Also, any change in sales growth and employee growth might be associated with variability in innovation, risk taking and firm’s age.


Introduction
Small and medium enterprises (SMEs) in general, they have more positive effect if they employ entrepreneurial orientation (Lee & Lim, 2009).SMEs drive a lot attention in the business world today and they can have different growth and development which reflects their capability of creating more or less jobs in the market.This paper addresses application of the EO dimensions in fast and slow growing companies -"gazelles and mices" (Birch, 1979) respectively, and its correlation with sales and employee growth.Why should we study these two types of SMEs?It is important because they contribute to economic development of a nation by creating most of jobs (Birch, 1979;Davidsson, P., L. Lindmark & C. Olofsson, 1994;1995;1996;1997;1997;OECD, 1987;Magnus & Johansson, 2010).Although these researches argue that gazelles create a "disproportionately" large number of new jobs comparing to mices, however, Neumark et al. (2008) have different view, arguing that mices have been classified as more jobs creators than it is reported.We believe that both of them are important and they are one of the most important pillars of economic growth of a state so we need to study them thoroughly.
Regarding the entrepreneurial orientation (EO), we can observe that for last three decades it has been studied using various measures for business performance of SMEs.A firm's upward growth could be related and connected to the entrepreneurship orientation (Brown et al., 2001;Stevenson & Jarillo, 1990).This relationship phenomenon of EO and firm's growth or performance is comprehensively researching from the theory (Covin & Slevin, 1991;Lumpkin & Dess, 1996) and practice (Covin & Slevin, 1989;Lumpkin & Dess, 2001;Wiklund & Shepherd, 2005).Despite this complete and extensive research there are some limitations and issues to be discussed.These are combined indicators, profitability and growth to associate with EO as sometimes contrary ones, and the developed latter concept of EO which is multidimensional (Moreno & Casillas, 2008).Entrepreneurial orientation has a positive effect towards a firm's performance and it helps the firm to grow (Wiklund & Shepherd, 2003).According to Kusumawardhani et al. (2009), he finds that SMEs with high level of EO have a better performance comparing to those who have a very low level of EO.However, on the other hand, Thomas and Mueller (2000) suggest that impact of EO to SMEs performance may differ from a country to a country.Depending to which part of the world a particular SME belongs it may happen that its adoption will vary.In this context Naldi et al. (2007) and Lee & Peterson (2000) argue that "national culture" affects the way how the EO could be applied.Thus, each culture of a nation should be carefully observed in order to apply EO as a positive effect to SMEs performance.This model is widely used to measure and compare business performance of the companies and organizations.
Studying both gazelles and mices is important due to the fact that they create a positive economic growth in terms of reduction in unemployment rate.Economic development of Bosnia and Herzegovina is an imperative and its economy needs at first place "gazelles" to improve economic outlook in general.In this content, mices should not be neglected as well.
The objective of this study is to examine a presence of EO dimensions in Bosnian gazelles and mices, its correlation with their business performance and what might be associated with change in their business performance (sales growth and employee growth).We examine application of the theory of EO in gazelles and mices developed by Covin and Slevin (1989), and which was further studied and developed by (Lee et al., 2001;Wiklund & Shepherd, 2005;Sue et al., 2011;Zhang & Zhang, 2012;Kraus et al., 2012;Kraus, 2013).
In what follows, section 2 briefly reviews EO and the EO dimensions.The section 3 links such literature to specific empirical analysis using a certain methodology.In section 4 and 5, we discuss obtained empirical results and derive important conclusions, managerial implications and limitations of the research.Why is this problem important?

EO Overview
The first concept of entrepreneurial orientation started with Miller (1983) who has depicted a company as entrepreneurial if ''one that engages in product market innovation, undertakes somewhat risky ventures, and is first to come up with proactive innovations, beating competitors to the punch''.This idea support Covin and Slevin (1989).This basic idea on entrepreneurial orientation is also adopted by others like Lee and Peterson (2000); Kreiser et al. (2002); Wiklund (1999), while Lumpkin and Dess (1996), have adapt the first idea from Miller (1983) and Covin and Slevin (1989) by adding two additional dimensions, i.e. autonomy and competitive aggressiveness.This concept is developed and discussed through decades and the following sections try to overview in a concise way recent work done by researchers on this topic.Miller (1983) argues that SMEs in order to grow and be sustainable it should be oriented towards the "product market innovation", which is followed by being a first in the market and ready to undertake the "risky ventures".Covin andSlevin (1989/1991) and Lumpkin and Dess (1996) suggest that an SME should not have administrative procedures that will prevent growth of the SME.In other words, the SME should have a leader with creativity that encourages innovation and eventually to "beat competitors to the punch" (Miller, 1983).His work supports Chow (2006) arguing that these three dimensions are unique and ample to affect SMEs performance, while Lumpkin and Dess (1996) and Hughes and Morgan (2007) argue that these additional dimensions, autonomy and competitor aggressiveness, are necessary because SMEs grow through different development stages.In a very similar way, Kreiser et al. (2002) argue that more options to improve SME's performance, the better chance for its improvement as additional two dimensions will provide.Lumpkin et al. (2009) confirm that the autonomy plays a strategic role in "in entrepreneurial value creation" and thus it should not be neglected.In the research done by Lee and Lim (2009), the EO proves its importance for services business.They argued that the EO is imperative to all SMEs' owners due to its positive effect.However, Su et al. (2011) find that there is a positive relationship between EO and firms' performance only for established firms and on the contrary it has inverse U-shaped curvature towards new ventures.Lee and Peterson (2000) study EO in the light of culture based on the Hofstede (1980) and Trompenaars (1994) cultural dimensions.They show that a culture in its nature is "low on power distance, weak in uncertainty avoidance and masculine, while a culture which is "individualistic, achievement oriented and universalistic" will generate efficient EO.This kind of EO will be attributed "by autonomy, proactiveness, competitive aggressiveness, innovativeness, and risk taking".Such powerful EO is able to pave the way to powerful entrepreneurship and globally recognized competitive advantage.As such, a brief overview of EO dimensions is followed by next subsections.

Innovativeness
Another term for innovation is innovativeness which, according to Lumpkin & Dess (1996), implies the propensity, creating new things, creativity in processes and experiments that leads to development of new products or services or even a new set of technological processes.Schumpeter (1942) gives a basis for this term arguing that innovative things combined together in the marketplace foster progress in particular society.
However, what might be a little confusing is that innovation or innovativeness is not the same as inventiveness.Inventiveness is a part of the general process of innovation and it is not enough per se for an SME to be innovative firm.It needed in SMEs but not enough for the SMEs prosperity and sustainability (Trevis et al., 2009).
Innovativeness in its essence implies an intension to back up new ideas and approaches, novel things, experiments, innovative procedures and steps starting with the built principles and technologies (Lumpkin and Dess, 1996).In the study done by Tajeddini (2010), a positive relationship between the Innovativeness and business performance in the services sector was found.

Proactiveness
What makes a firm proactive?The proactive firm is always first entrants into a new market and they are the first or pioneer in that particular business (Trevis et al., 2009).To be first in the market an organization needs to be "first-mover" in the market providing possible offerings to the market that is based on the needs of customers (Lumpkin & Dess, 1996).Once the firm has its advantage of being a "first-mover" then the firm has open the door to get advantage of this role by "skimming" pricing strategy possibly to be applied in such market (Zahra & Covin, 1995).Similarly, it can become a market controller in monitoring "distributing chains" and it may welcome to establish "brand recognition" very quickly (Wiklund & Shepherd, 2003).

Risk-Taking
Risk taking is related to the risk of turnover or to the risk of other financial analysis (Trevis et al., 2009) and anything that is huge action of let say, "borrowing" in capital and financial terms that might face the uncertainty could be regarded as risk taking (Baird & Thomas, 1985).The risk could be perceived as a company's intention to be involved in lofty risk projects and managerial options opposite to prudent actions (Miller, 1983).
One may pose a question how to define when an entrepreneur is risk taker?It is complex since entrepreneurs do perceive these things differently.Recent research has argued that from entrepreneurs' perspective, their actions are not risky (Simon, Houghton, & Aquino, 2000), and only action, to reduce the risk, is undertaken after planning and anticipating all circumstances (Bhide, 2000).

Entrepreneurial Orientation (EO) Development and Performance
As everything, the EO is evolving from time to time and applied in a different ways and to different organizations types.
The very beginning of EO started with Schumpeter (1942) who stresses out the importance of entrepreneurial orientation.The table 1 shows the EO chronological developments that lasted over three decades.
In regard of business performance that could be applied in the research of SME and EO related there are many measures that have been undertook.The "performance" itself might be measured in perceived financial terms, perceived non-financial and "archivical financials" or eventually to make any of possible combinations in the particular research (Rauch et al., 2009).In our study, similar to Kraus et al. (2012), we take perceived financial measures due to unavailability financial data of the companies in Bosnia and Herzegovina.Bamford et al. (2000, p. 255) consider this data as disadvantageous that may negatively affect results reliability.Others found that perceived financial rates are more less the same and in many cases.
Research undertaken by Lee and Lim (2009), show a positive relationships between EO and business performance of the services firms arguing that EO development among the SMEs' owners is a receipts for their growth and development.Note.Overall performance is "perceived and evaluated by the small business owner (and not performance as defined by return on sales -ROS or return on investments -ROI or any other specific measure whose very meaning may depend on EO and/or SBO)" (Runyan et al., 2008).Source: Author's compilation.
This table shows a relationship between EO dimensions and business performances (with different measurements) of SMEs with different business orientation (manufacturing, services, technology, retail/wholesale, and others).Most of them have been proved that EO dimensions had a positive relationship with business performance of firms.However, one of them has found a negative relationship between EO dimensions and business performance of firms (Slater & Narver, 2000).

Data
In this research we have used a self-administered questionnaire to collect data based on Covin and Slevin (1989) model.Its original version has been translated into Bosnian language and then reversed back to English.There were no inconsistencies that may affect the original model so it is preceded with Bosnian version of questionnaire.So the model of EO dimensions from Covin and Slevin (1989) is adopted and modified for the research.The 7-Likert opposite scale has been used, where 1 is strongly disagree and 7 is strongly agree.
In order to identify gazelles and mices we use Davidsson and Henrekson (2002) applying annual sales growth of 5 or 10%, and Hölzl, W. ( 2009), seeing gazelles as companies whose annual employee growth are 10% with inconsistencies of 5%.
Stratified sampling has been used.From the database of SMEs provided by Federal Ministry for Entrepreneurship, Development and Turnover and Foreign Trade Chamber of Bosnia and Herzegovina, which contain 16,480 SMEs, we randomly selected 5% (in total 824 SMEs).We received 205 responses and 27 were rejected due to our stratification strategy that SMEs were identified as either gazelles or mices, we end with 178 SMEs which corresponds to 21, 6%.So, the effective sample size was 178 split up into two stratums, gazelles (89) and mices (89).

Factor Analysis
We use factor analysis to check whether the data are fully loaded and to measure construct validity and its multidimensionality (Nunnally, 1978, Kraus et al., 2012).All independent variables, the EO dimensions, were loaded to perform these measurements.Principal component analysis (PCA) is applied to identify a path or a few paths that could explain their variability as much as possible (Belle et al., 2004, p. 589).We apply orthogonal and Varimax method.Orthogonal (unrelated) rotation is rotating factors with keeping all variables independent.Varimax method is good to minimize dispersion of loaded items within factors so that a smaller numbers of loaded variables will be loaded in a very high manner (Field, 2005, pp. 635-637).To determine number of loading factors we used the Kaiser Criterion that is set out as eigenvalues>1 (Kaiser, 1960).Factor loadings results may depend on sample size.Sevens (1992) argues that a sample size larger than 100 is to be as appropriate to get acceptable results as of 0.52.Kaiser (1974) argues that any value greater than 0.5 is acceptable.
The results have shown that we have three components and we maximized loadings on each variable of extracted factors while minimized loadings are on all other factors.Initial eigenvalues and extracted percentage of explained variances are the same, whereas factor 1 has higher value (42,039%) comparing to the rest two (17, 69% and 11, 55% respectively).After rotation of variables, the total % of explained variance was for factor 1 29, 79%, factor 2 29, 23% and 12, 26% of the factor 3 (See Table 4).Bertlett's test of sphericity applies control whether the PCA has a point to be set out and if p-value is less than 0.001.In our case it is significant (Sig=.000).The KMO measure was obtained as 0.801, with Chi-square of 638.035, (df=36), which is more than required (>0.5, Kaiser, 1974), and it has proved sampling adequacy.The highest score of EO dimensions in gazelles is for innovation, while the least one is for proactiveness.In mices firms, the highest score is from proactiveness too, while the poorest score is on risk taking dimension as well (Table 5).To confirm the previous result and get decision rule (if p<0.05, reject H 0 ) on H0 and H 1 , we use the t-test.The t-values with its degree of freedom (df=176) for two EO dimensions are high for two predictors resulting in its significance, Sig=0.001;0.007 respectively, (innovation, risk-taking) confirmed that H0 is rejected, confirming H 1 (H 1a and H 1c ).Moreover, although the total score of mean for proactiveness is higher in comparison to one in mices, the t-test do not reject and H0b (p>0.05) and H 1b is not supported (Table 6).
Correlation matrix has shown the possible correlation between predicted and criterion variables.In this case predicted variables are innovation, proactiveness, risk taking while criterion variables (dependent) are sales and employee growth over last 3 years or less if a firm exists less than 3 years.Results are reported in the Table 7. Proactiveness, among the rest EO dimensions, has no positive correlation with sales growth (0.057) but not significant (p>0.05).The same is true for correlation between proactiveness and employee growth (0.061; p>0.05).Innovation has significantly positive correlation with sales growth (almost moderate correlation with 0.250, p<0.001) while its correlation with employee growth have also moderate correlation value of 0.296 and yet significant (p<0.001).Risk-taking has lower correlation coefficients with sales growth (0.203; p<0.007) and employee growth (0.228; p<0.002).So, H 1d is partially supported (innovation and risk taking have positive correlation with sales and employee growth).These results reject H 0d and support H 1d. Since we have partially confirmed hypotheses that we tested, we want to see how much of predictor variables could be explained in criterion variables.We use linear regression to create a combination that will fit the regression model.In this case we had 2 criterion variables, sales and employee growth and three predicted variables, innovation, proactiveness and risk-taking.Controlling variable is firm age of firms.

Criterion 1: Sales growth
Linear regression was used to examine whether EO dimensions (innovation, proactiveness, risk-taking) and together with control variables (firm type, firm age and firm size) significantly predict relationship with sales growth and employee growth rate over last three years.Two models were setup.In the first model we include innovation, proactiveness and risk-taking as independent variables.In the second model we add firm age.Firm age was added to second model.
The null hypothesis: H 0d = β 1 = β 2 = β 3 = β 4 ; EO dimensions do not have a significant relationship with business performance and growth of gazelles and mices in Bosnia and Herzegovina Alternative hypotheses: H 1d : β j ≠ 0; for at least one of j, where j=1, 2, 3… n; EO dimensions have a significant relationship with business performance and growth of gazelles and mices in Bosnia and Herzegovina.
First model (Table 8) made of predictor variables, innovation, proactiveness and risk taking we have got R² change of 0.096.So 9.6% variability has been counted by these EO dimensions.For model 2, along with EO dimensions, we added firm age as additional predictor variable, and results have shown that there is a slight increase in R²=0.136 with a change of 0.040 or 4% of its variability.The ANOVA table (Table 9) results shown that the all variables in its aggregate for 2 models were significant predictors (independent variables along with control variable) of sales growth, whereas explaining 8.1% of the variation in the model 1,F (3, 174) = 6.176, p = .001.In the model 2, they explain 11.6% of their variability, F (4, 173) = 8.027, p = .005.These results imply rejection of the null hypothesis that says "EO dimensions do not have a significant relationship with business performance and growth of gazelles and mices in Bosnia and Herzegovina".
Results of linear regression models are shown in the table 11.
The following equation presents the general linear regression model: where, Y i is criterion variable (dependent variable), β 0 is constant where the Y axis is intercepted by regression line; β 1 and β n are the coefficients of the slope for regression predictors; X i1 and X n are the values of predictors values (independent variables) for some of i'th observation; ε i is residual.
In this context we have model 1: Model 2: In terms of contribution of each predictor variable as well as control variable that is examined in 2 models, it shows that in two models, innovation, risk taking and the firm age are significant.Only, EO proactiveness is not significant.So independent variables (innovation and risk-taking) along with control variable (firm age), make a significant contribution to the models and therefore they were significant predictors of sales growth (Table10).
These results imply rejection of the null hypothesis that says "EO dimensions do not have a significant relationship with business performance and growth of gazelles and mices in Bosnia and Herzegovina".The ANOVA (Table 12) results shown that overall 2 models were significant predictors of employee growth, whereas explaining 11.7% of the variance in the model 1,F (3, 174) = 8.849, p = .000.In the model 2, they explain 12.5% of their variability, F (4, 173) = 7.344, p = .000.The table 13 shows coefficients of the two models of the criterion 2 (employee growth) with their corresponding values of t.The two models with EO dimensions have contribution to variability in change of employee growth, while adding the control variable firm age, does not contribute to its change as it is shown in the model 2.
The following table (Table 14) reviews developed hypotheses and their outcomes.
Table 14.Hypotheses testing results

Hypotheses Results
H1. Entrepreneurial orientation dimensions are more present in "gazelles" companies while less present in "mices" companies and EO dimensions have a significant relationship with their business performance and growth in Bosnia and Herzegovina.

Partially supported
H1a. Innovation is more present in "gazelles" while it is less present in "mices".Supported H1b. Proactiveness is more present in "gazelles" while it is less present in "mices".

Not supported
H1c. Risk taking is more present in "gazelles" companies while it is less present in "mices".Supported H1d. EO dimensions have a significant correlation and relationship with business performance and growth of gazelles and mices in Bosnia and Herzegovina.Supported

Conclusion
The research on gazelles and mices (Birch, 1979) is done on EO theory framework developed by Covin and Slevin (1989).It analyses several hypotheses.First hypothesis states that "Entrepreneurial orientation dimensions are more present in "gazelles" companies while less present in "mices" companies and EO dimensions have a significant relationship with their business performance and growth in Bosnia and Herzegovina".Empirical results have shown that EO dimensions are more present in gazelles than in mices except proactivenes.This might be due to unwillingness of business owners and managers to take their moves ahead comparing to others.It is been proved that gazelles had more weight in terms of EO dimension comparing to mices (See Table 6).Correlation analysis shown significant low to moderate correlations between independent variables (innovation and risk taking) and dependent variables (sales growth and employee growth), while proactiveness did not show any correlation towards sales growth and employee growth.Similarly, in overall the "EO dimensions along with firms' age factor contribute to changes in sales and employee growth of both gazelles and mices" (Table 14).
In criterion 1, the sales growth, among predictor variables, only proactiveness has not been observed as significant in both models, while innovation and risk taking, accompanied with firm age, predicted variability and change in sales growth.In criterion 2, it was observed that EO dimensions had contributed to a change of employee growth.However, the firm age, has not bring any contribution in change of the employee growth.
Our results suggest that SMEs which want to have a significant business growth and develop faster (to become gazelles) they should incorporate innovation, proactiveness and risk taking.In addition, firm age of a SME gives more opportunity to achieve better business performance results.Meaning that, business owners and managers should look forward to be innovative, and when necessary risk-taking roles in the company.
Due to reluctance of the participants this research is lacking of a decent sample size (n=178).In addition, achieved data for dependent variables, are known as perceived financial data that cannot predict eventual causality between dependent and independent variables.Also, "perceived financials", might be too subjective which also may reflect in the end results.Aforementioned limitations should be first task done for the same research.Also, we strongly suggest qualitative data that will support or reject objectives and hypotheses.

Table 1 .
Summary of EO development Source: Author's compilation.

Table 2 .
Summary of overviewed EO dimensions and relationship with business performance Kraus et al (2012)1): It is positive for established firms and not for new ventures.Kraus et al (2012): Positive only for proactiveness, innovativeness and risk taking has shown no positive relationships.Source: Author's compilation.

Table 3 .
Summary of SMEs' business measures, sample size and industry Efficiency (return on investment, return on equity, and return on assets), growth (sale growth, employee growth, and market share growth), and profit (return on sales, net profit margin, and gross profit margin).

Table 4 .
Total variance explained for eigenvales, extraction and rotation loadings

Table 5 .
Comparison of means for gazelles and mices

Table 7 .
Correlation matrix results

Table 8 .
Regression model summary of results of 4 models: criterion 1 (sales growth) c. Dependent variable: sales growth.

Table 9 .
ANOVA results for criterion 1 sales growth

Table 10 .
Beta coefficients and t-statistics summary: criterion 1 (sales growth) Beta coefficients and t-statistics table(Table 10) show that prediction of criterion variable could be found in the model 1 and model 2. First model shows its significance in contribution of all three EO dimensions to employee growth.When we add control variable, the firm age, to EO dimensions, it does not significantly contribute to change in employee growth.On the contrary, other independent variables (innovation, proactiveness and risk-taking) show significance for variability in employee growth.Results of the second criterion, employee growth, are shown in the table 11.Two models we develop and in the first model are the following predictor variables, innovation, proactiveness and risk taking.The R² (0.132), 13.2% of has been counted by these EO dimensions.For model 2 we add to EO dimensions the firm age as additional predictor variable, and results have shown that there is a little change in R² (0.145) of 0.013% (1, 3%).

Table 12 .
ANOVA results for criterion 2 employee growth

Table 13 .
Beta coefficients and t-statistics summary: criterion 2 (employee growth)