The Development of MSMEs International Networks through Internationalization

This article studies the effects of relationship bonding, entrepreneurial cognition, and market dynamism towards internationalization and their impacts in MSMEs network growth. This article examines hypotheses derived from theory and research employing structural model analysis in 150 exporting-MSMEs in Central Java province, Indonesia. This research reveals that relationship bonding influences internationalization while entrepreneurial cognition and market dynamism do not. Internationalization affects the MSMEs network growth. This leads to the finding that MSMEs internationalization in Central Java, Indonesia is promoted more by social factors instead of individual and macro environment, while network growth is stimulated by individual and macro environment instead of social factors. Social determinant directs internationalization and individual and macro determinant regulates network expansion. The implementation of network theory and ecosystem management theory to comprehend MSMEs internationalization and network growth. This research provides supports to encourage educators and policy makers upgrade the skills of MSMEs practitioners in developing international business. This research compares social perspective, individual perspective, and macro environment in an empirical study in a developing country. This research offers solution to comprehend internationalization and network growth involving market dynamism and entrepreneurial cognition as well as bonding relationship as determinants.


INTRODUCTION
Internet development benefits MSMEs in marketing their products overseas. Approximately 75% of economic impact provided by internet is taken up by MSMEs traditional companies that only sell their products and services on the internet by employing its high productivity (Manyika and Roxburgh, 2011). The MSMEs vast development in a country is essential as MSMEs are tools to build entrepreneurship, innovation route, industrial dynamic source, and work creator instrument (Ahmed, 2015). These are the reasons for broader intensive MSMEs development. This development can be achieved through internationalization. Internationalization is state boundaries crossing in an organization"s growth process (Child and Rodrigues, 2005). There are many variables proven to influence MSMEs internationalization. They are grouped into strategic variables (like opportunities) (Dunning, 2001), economic variables (Ghoshal, 1987), and non-economic variables (George, Wiklund and Zahra, 2005).
Nevertheless, the previous studies on non-economic factor are divided between whether MSMEs internationalization is influenced by cognitive or behavioral factor only, and whether it is also influenced by social and individual factor (Dew, Grichnik, Mayer-haug, Read, and Brinckmann, 2014;Rae, 2005). Those are still unanswered and require further studies. Some research have attempted to comprehensively unravel the question, but indirectly employing cognitive construct (Kyvik, Saris, Bonet, and Felício, 2013;Zahra, Korri, and Yu, 2005) and behavioral construct (Moen and Servais, 2002;Reihlen and Apel, 2007). Research involving individual factors (Zahra et al, 2005;Kyvik et al, 2013) and social factors (Ellis, 2011;Arenius, 2002;Chetty and Agndal, 2006) have also been conducted, although without comprehensive construct such as entrepreneurial cognition representing individual and social bonding factors as collective factors. Therefore, more comprehensive and inclusive research are needed when testing factors influencing internationalization and MSMEs network development. Furthermore, internationalization theory has not been broadened to consider three things at once: social factors, rational factors, and contingency factors. As far as the writers know, these come from three bigger theories namely network development (Johanson and Mattson, 1987), FDI (Foreign Direct Investment) theory (Caves, 1971), and dynamic capability theory (Teece, Pisano, and Shuen, 1997). Employing structural survey, this article aims to explore the correlation of bonding, entrepreneurial cognition, market dynamism, internationalization, and network development in MSMEs that have internationalized in Central Java province, Indonesia. There are three contributions emanated from this research. First, this research contributes in internationalization literature by showing how variables focusing on individuals, groups, and environment affect MSMEs internationalization work. This research is also instrumental in literature by providing empirical validation for the correlation of social factors, rational factors, and contingency factors in internationalization. Second, this research operates network theory in bonding variable, FDI theory in entrepreneurial cognition, and dynamic capability in market dynamism. Third, based on Indonesian internationalized MSMEs samples, this research explores the effects of those variables to internationalization and network development, which leads to validating findings obtained in industrialized countries (for example Knight and Liesch, 2016;Vatamanescu, Zbuchea, Pinzaru, and Andrei, 2016;Veugelers, 2008). This eventually provides description of internationalization behavior in developing countries. It is expected that this research can act as a base for future research exploring MSMEs internationalization, especially in developing countries.

THEORETICAL FRAMEWORK
The social aspect of entrepreneurial construct related to internationalization generally employs social capital variable (Arenius, 2002;Chetty and Agndal, 2007;Ellis, 2011). It is generally understood that good social capital generates positive effect to internationalization. One form of social capital is bonding. Bonding is a behavioral variable as it is derived from the application of some techniques such as customization, information sharing, and managerial bonding development (Davcik and Sharma, 2016). Bonding can also be a collective variable because it involves perspective of an existing social network. Literature in marketing classify three types of bonding namely economic bonding, social bonding, and structural bonding (Harrison-walker and Neeley, 2012). Economic bonding is the amount of money and time spent to build bonding, social bonding is virtual interactions among the members to build connections, and structural bonding is commitment given to the network that makes it difficult to leave the bonding (Quinton and Harridge-march, 2010). The bonding applied in this research is social bonding, to make it in line with social capital theory. Bonding capital is a part of social capital that better suits MSMEs, as MSMEs have relatively weak bargain position that forces them to rely on bonding mechanism in aligning interest in behavioral-based contract (Arenius, 2002). Nevertheless, there is only one research investigating the correlation of bonding capital and internationalization . The research reveals that bonding capital affects positively towards MSMEs internationalization. Overall literatures support the notion that social capital positively correlates to internationalization speed (Arenius, 2002). Apart from using social capital theory, bonding can also be analyzed by network theory. In accordance with network theory, internationalization is strongly affected by network of formal and informal bonding (Coviello and Martin, 1999). Bonding is a form of formal and informal network meant by network theory. Thus, the following hypothesis is proposed: Hypothesis 1: Bonding will show positive correlation with MSMEs internationalization Hypothesis 2: Bonding will show positive correlation with MSMEs network development. The concept of entrepreneurial cognition is a new one developing in entrepreneurship studies (Mitchell et al., 2002). Studies in entrepreneurial cognition have emerged in 1976, but this was still implicit in nature that it faced conceptual challenge (Grégoire, Corbett, and Mcmullen, 2011). Phenomena as learning failure and problem solving in business are clearly cognitive phenomena in entrepreneurship (Corbett, Neck, and Detienne, 2007). Other similar things are how one learns how to become an entrepreneur and correlates his knowledge with his business intention (Pihie, Bagheri, and Sani, 2013). However, the concept of entrepreneurial cognition has just been developed to include all cognitive aspects that have been learned in the previous entrepreneurship studies. A basic question in entrepreneurial cognition is how an entrepreneur thinks (Tang, Kacmar, and Busenitz, 2012). In this case, several constructs have been proposed like entrepreneurial perception, entrepreneurial intention, entrepreneurial efficacy, faith in entrepreneurship, entrepreneurship study, entrepreneurial capacity, and so forth (Acs and Audretsch, 2010;McKenzie, Ugbah, and Smothers, 2007). It is still in debate whether this cognition is formed individually or collectively (Dew et al., 2014) and whether cognition, without including behavior, can lead to internationalization (Rae, 2005). Research in entrepreneurial cognition have revealed that there is correlation between entrepreneurial cognition and internationalization (Knight and Liesch, 2016). Entrepreneurial cognition is regarded as how an entrepreneur proactively takes steps in internationalization, especially one who has previously had experience in international business be it personally or as an employee in multinational company (Aspelund, Madsen, and Moen, 2007;Luostarinen and Gabrielsson, 2006;Mcdougall, Oviatt, and Schrader, 2003;Weerawardena, Sullivan, Liesch, and Knight, 2007). Entrepreneurial cognition can be seen as individual characteristic determining if an MSME takes internationalization step from the beginning (Aspelund et al., 2007;Luostarinen and Gabrielsson, 2006;Mcdougall et al., 2003). Apart from experience, specialized knowledge seems to play role in assisting internationalization (Fan and Phan, 2014;Rialp, Rialp, and Knight, 2005;Weerawardena et al., 2007). Learning orientation, as one cognitive aspects, has also known to affect as internationalization booster (Baum, Schwens, and Kabst, 2011;Freeman, Edward, and Schroder, 2006;Kuivalainen, Sundqvist, and Servais, 2007;Mudambi and Zahra, 2007;Rialp et al., 2005). Entrepreneurial cognition should be seen as capability in supporting internationalization (Rialp et al., 2005). As it is recognized, capability and competence are parts of performance determinants, superior in resource-based view perspective. Moreover, entrepreneurial cognition provides superior market knowledge for entrepreneur, which eventually gives benefits triggering internationalization (Aspelund et al., 2007;Fan and Phan, 2014;Freeman and Cavusgil, 2007;Garcia-Canal, Duarte, Criado, and Llaneza, 2002;Mathews and Zander, 2007;Mcgaughey, 2007;Mudambi and Zahra, 2007;Rialp et al., 2005). Manager with entrepreneurial cognition will be able to understand, recognize, and exploit opportunities existing in international market (Zahra et al., 2005). This ability is highly necessary as rushing internationalization will ruin MSMEs performance (Mudambi and Zahra, 2007). However, some research have indicated that early internationalization does increase capability that boosts MSMEs performance (Autio, Sapienza, and Almeida, 2000;Zahra, Ireland, and Hitt, 2000;L. Zhou, Barnes, and Lu, 2010). The correlation of entrepreneurial cognition and internationalization can be elaborated by employing FDI theory (Coviello and Martin, 1999). According to this theory, internationalization is a result of rational decision made by entrepreneur in developing his business. Often times this is a gradual decision in which company first runs in local market, then moves to international one. The research in entrepreneurial orientation reveals that this concept positively affects MSMEs development (Wiklund, 1998). A study conducted by Wiklund and Shepherd (2005) found that environmental dynamic and entrepreneurial orientation have roles in promoting MSMEs development. This is in line with resourcebased view that human resource is the most important resource in MSMEs development (Wiklund, Patzelt, and Shepherd, 2009). It can be concluded then, from entrepreneurial cognition studies and FDI theory mentioned earlier, that entrepreneurial cognition will affect MSMEs internationalization. Thus, the following hypotheses are proposed: Hypothesis 3: Entrepreneurial cognition will show positive correlation with MSMEs internationalization Hypothesis 4: Entrepreneurial cognition will show positive correlation with MSMEs network development Market dynamism represents contingency factor which a determinant in MSMEs internationalization and network development. This factor is employed in dynamic capability theoretical model (Atanassova and Clark, 2015), in which a company"s dynamic capability is determined by market dynamism (Eisenhardt and Martin, 2000). Dynamic capability is organizational and managerial process of creating, coordinating, integrating, reconfiguring, and altering resources and skill to maintain business and increase competitiveness by adapting to environmental changes (Atanassova and Clark, 2015;Eisenhardt and Martin, 2000;Teece, 2007;Wang and Ahmed, 2007). Competition creates market dynamism (Stone and Badawy, 2011). Market dynamism is among of external triggers to generate dynamic capability of an organization (Atanassova and Clark, 2015;Cavalcante, Kesting, and Ulhøi, 2011;Du and Banwo, 2015). It is generally known that market changes create more growth opportunities (Wiklund et al., 2009;H. Zhou and de Witt, 2009). The growth opportunity is utilized by MSMEs to introduce new product/service in the market, therefore prompting competitive advantage and yielding better financial result (Wiklund, 1998). This opportunity is also used to expand the market to sell the products and services and therefore creating new networks in many places (Ratten and Dana, 2007). Thus, market dynamism is found in various research as a growth determinant for high profile company (Moncada-paternò-castello and Cincera, 2012; Veugelers, 2008). Based on vast accepted correlation between market dynamism and MSMEs development, we presume that: Hypothesis 5: Market dynamism will show positive correlation with MSMEs internationalization Hypothesis 6: Market dynamism will show positive dynamic with MSMEs network development Unlike FDI theory, the network theory holds that social bonding contributes role and large influence towards business transactions in internationalization (Johanson and Mattsson, 1987). This theory opposed FDI theory that states internationalization as of rational strategic decision making process (Coviello and Martin, 1999). According to network theory, internationalization is the result of members" behavior pattern inside the network that creates formal and informal networks. Bonding interaction and development establish internationalization of a business. The company in time uses the available network bonding to proactively execute its business (Boojihawon, 2004). Network theory has been applied in bonding variable. The yielded social capital, as a source of competitive advantage, should be developed to generate more capital. This means that when internationalization has been achieved, there will be more efforts to widen the networks into more areas. Therefore, MSMEs develop into international market. This has led to the following hypotheses: Hypothesis 7: MSMEs internationalization will show positive correlation with MSMEs network development

METHODS
The population of this research is MSMEs in Central Java province that have exported their products. There are 973 MSMEs involve in this research. Their data are obtained from Dinas Koperasi dan UMKM (Cooperative and MSMEs Service) Central Java. Samples are determined by considering the minimum required sample in SEM (Structural Equation Modeling) analysis, which is the data analysis method applied in this research. Hair, Eingle and Sarstedt (2011) state that minimum sample of SEM analysis is ten times the questions in the variable with the most questions. Variable with the most questions here is entrepreneurial cognition variable, consisting of 20 questions. Hence the minimum sample is 20 x 10 = 200. The researchers managed to obtain 150 valid questionnaire results. Although this seems low, we argue that our samples represent 15,4% of the total exporting-MSMEs population in the location where the study is conducted. This means that the number is quite representative. The detailed samples characteristics are shown in Table 1. Data are collected in three different techniques namely email (if the MSMEs have it), direct field visits, and fellow researchers" assistance. When responses have reached 200 or when the entire population has filled out the questionnaires, we begin the data analysis. Different test is carried out to examine the way three methods fill out the questionnaires in order to seek for potential bias. If significant different results present, data collecting method variable needs to be controlled by incorporating it as variable in analyzing the data.
Bonding relationship is measured by using instrument comprises of three related items in connection to MSMEs and their overseas partners. The instrument is adapted from Woodside and Baxter (2015). Those items are strong social bonding among the people involved in business partners" organizations overseas, highly advantageous bonding, and the many shared goals possessed by MSMEs and their overseas partners. They have a 5option scale ranging from "strongly disagree" (coded 1) to "strongly agree" (coded 5).
Entrepreneurial cognition employs questionnaires by Mitchell, Smith, Seawright, and Morse (2000) developed by Urban (2008), which consist of willingness dimension and ability dimension. Total 20items are used to measure entrepreneurial cognition in this research. They are valued in binary option scale. The first option indicates high entrepreneurial cognition while the second one does not. The more options which are made for the first option means that the higher the entrepreneurial cognition is.
Market dynamism is a two-indicator variable. The specific items applied are adapted from Zhou and de Wit (2009). Dynamic market indicators are measured in 5-point Likert scale, where "1" means "strongly disagree" and "5" means "strongly agree." Meanwhile internationalization indicators are valued in the same 5-point Likert scale but with different options for each item which generally provides composite view for the option taken (for example 1 means 1-20%; 2 means 21-40%, and so on).
MSMEs internationalization is measured in four indicators, including (i) percentage of foreign market profits, (ii) percentage of sales revenue of foreign market, (iii) the number of employees working for foreign market, and (iv) the ability of organization structure to explicitly accommodates international activities (Asugman, Johnson, and Mccullough, 1997).
Meanwhile international network utilizes instrument of Mudambi and Zahra (2007). MSMEs network development variable is measured in three items: (i) the number of export destination country growth, (ii) the number of companies cooperating with the MSMEs growth, and (iii) MSMEs export intensity growth.
Respondents are also asked to provide information related to age (under 35 years old; between 35-55 years old; more than 55 years old); the number of foreign partners (1-10; 11-20; 21-30; 31-40; more than 40); education (elementary school; junior high school; high school/vocational school; diploma; graduate; post graduate); field of study (economics; others); and experience in international business (1-3 years, 4-7 years, 8-10 years, more than 10 years). Data analysis of this research is carried out by employing multivariate statistical techniques Structural Equation Modeling (SEM).

RESULT
Confirmatory analysis is conducted to check the reliability of all the indicators using their loading. As good reliability and validity are hard to reach for entrepreneurial cognition variable, it is reduced into three items with the biggest factor loads. The same thing happens to market dynamism variable, which is not reliable. Since it contains only two items, we separate this variable into two single items. Internationalization variable is also reduced into one item out of four previously planned. Relationship bonding and growth variables are not reduced because they are valid and reliable. Table 2 shows that all indicators have reached the accepted value as they have passed the defined threshold. VIF values of each construct are less than 5, indicating the absence of multicollinearity problem (see Table 2).   This research examines seven hypotheses to predict MSMEs internationalization and growth by employing relationship bonding variable, entrepreneurial cognition variable, and market dynamism variable. In line with Figure 1, the result accepts hypothesis 1 and hypothesis 7. Hypothesis 4 and 6 are relatively mixed due to two accounted variables: entrepreneurial cognition and market dynamism, which each consists of two dimensions and only one dimension shows relationship in the hypothesized direction. Meanwhile the result rejects hypothesis 2, 3, and 5.
Bonding relationship, entrepreneurial cognition, and market dynamism variables are proven to predict 6,9% variation in MSMEs internationalization. Meanwhile these variables together with internationalization predict 30,6% variant of network growth.
To have knowledge on the model"s predictive validity, we calculate the value of Stone-Geisser Q2. A less than 0 Q2 value indicates that the model has relevant predictive value (Hair Jr, Hult, Ringle, and Sarstedt, 2016). The Q2 value of this model is 0,354 and thus proves big predictive relevance.

DISCUSSION
For MSMEs samples in Central Java, Indonesia, the relationship bonding variable is the strongest variable influencing internationalization. On the other hand, entrepreneurial cognition and market dynamism variable play roles in network growth. The insignificant correlation between relationship bonding and network growth does not accord to the view that social capital will speed up internationalization (Arenius, 2002). Nevertheless, relationship bonding does create positive correlation to internationalization, as revealed by Vatamanescu et al (2016). The plausible explanation of the insignificant effect of relationship bonding towards network development is that MSMEs operate in close network which focuses only on the existing relationships (Behari, 2016). Since the consumers" focus is on this, the network development will depend on the consumer-triggered growth instead of on the MSMEs" themselves. There has been internal growth in the existing network which happens qualitatively. This confirms the finding in management study where tight relationship bonding results in mutually relationship of the parties involved (Nuryakin, Aryanto, and Setiawan, 2018). That is worth seeking as it promises large advantages in business relationship (Paduraru et al., 2016). In the meantime, when network growth is on the go, a commitment needs to be made which means that relationship bonding should be developed with new partners. There is a need for cultural investment (Billore and Borg, 2014) in it and this does not depend on the relationship with the existing partner. This is related more to the internationalization, entrepreneurial cognition, and market dynamism, which show significant correlation in network growth in this research. It will be difficult for MSMEs to manage complex network in line with the development of their networks (Boojihawon, 2004), while the existing relationship bonding has guaranteed well growth. This research has also found that internationalization is only significantly influenced by relationship bonding which is a social factor. Cognitive variable in terms of entrepreneurial cognition does not affect internationalization. Market dynamism variable, which is a contingency variable, does not correlate significantly to internationalization. This means that of the three factors: social, rational, and contingency, only the social factor has been proven to promote MSMEs internationalization, at least in MSMEs sample in Central Java. This finding reaffirms social perspective of MSMEs internationalization (Ellis, 2011;Arenius, 2002;Chetty and Agndal, 2006). Although market dynamism does not affect internationalization, it does influence the network growth. However, the two variables employed yield different results. Product change variable negatively affects the network development, unlike customer needs variable that positively influences it. The reasonable clarification of the negative effect towards market dynamism in product change dimension is that product change disrupts the existing pattern of MSMEs products. This in turn will discourage MSMEs to develop their networks. Research conducted in Indonesia also reveal that MSMEs prefer to maintain the existing profit instead of do expansion (Ikhsan, Almahendra, and Budiarto, 2017). MSMEs respond by reducing customers and focusing on costumers who demand their products in traditional form which is their specialty. In other words, MSMEs are unable to cope with product change requested by dynamic market. This confirms the finding that the bigger the company is, the more adaptive it is to cope with dynamic environment (Yu, Lumpkin, Parboteeah, and Stambaugh, 2019). The research shows that highly dynamic environment may negatively influence company"s performance (Davis, Eisenhardt, and Bingham, 2009). Product change is different from customers" taste change, because customers might want something which is not totally new -a variation of the old product. This is why the other variable of market dynamism, namely customer needs, positively and significantly influences network development.

CONCLUSION AND IMPLICATION
Overall, the result of this research supports the network theory because relationship bonding significantly influences internationalization. Nevertheless, since international network growth can also be seen as internationalization variable, supports are also provided for FDI theory and dynamic capability theory owing to the finding that entrepreneurial cognition and market dynamism affect the MSMEs" international network growth. This finding has important implication both theoretically and practically. First, this research adds to the existing literature that market dynamism contributes in MSMEs development in complex manner (Davis et al, 2009). Dynamic market does not necessarily encourage MSMEs development as the MSMEs may opt to hold out, withdraw, or follow the changes by innovating. It means that market dynamism is a strategic bait for MSMEs (Seo and Chae, 2016). This explains why this research as well as other research in developing countries (Donkor, Donkor, and Kwarteng, 2018) find that market dynamism will not affect MSMEs performances (Donkor et al, 2018). This can be assessed from ecosystem management theory perspective which affirms the strategic role of market dynamism in shaping internal dynamic of a company to manage their resources (Crielaard, Wubben, and Omta, 2018). The second theoretical implication relates to theoretical contestation in MSMEs internationalization. Since internationalization in this research is influenced solely by relationship bonding, the finding supports the network theory more (Coviello and Martin, 1999), than FDI theory (Caves, 1971) or dynamic capability theory (Teece et al, 1997). Eventually, empirical validation can only be provided for social factor and not for rational factor of FDI perspective and contingency factor of dynamic capability perspective. Third, this research encourages review towards entrepreneurial cognition. It shows that only ability aspect can promote international entrepreneurial activities. Willingness aspect alone cannot generate either internationalization or international network growth. This means that sharp theoretical difference is necessary to separate willingness from ability as two different constructs in entrepreneurial studies. The gap between willingness and ability is a common thing found in economic practice (Ashraf, Glaeser, and Ponzetto, 2016). Some entrepreneurs have willingness but have no ability, while the other have the ability but without willingness (Richards, Kammerlander, and Zellweger, 2019). In this case, entrepreneurial cognition as a variable with willingness and ability dimension might find it hard to produce valid construct. Entrepreneurial cognition variable can only be formed as a unity when the motivation gap separating willingness and ability (Kotlar, de Massis, Frattini, and Kammerlander, 2016) can be closed.
In the meantime, the practical implication emerged is that people of Central Java should build a belief that they are able to develop their MSMEs businesses to international level. Just willingness is not enough to create MSMEs business growth. This research reveals that it is the ability script, not the willingness script, which creates MSMEs network growth internationally. To develop ability, one needs to take concrete steps by expanding experience or by taking training and education. Therefore, MSMEs administrators should spare their times to join entrepreneurial trainings held by government, especially the ones focus on ability development, instead of just be aware of it. This kind of training helps to develop their ability in effectively handling international business (Vahidnia et al., 2019). This means that other stakeholders like educators and policy makers should also develop training and education program that promote MSMEs practitioners" learning activity (Keane, 2016). These trainings and educations have been implemented in various ministry programs such as PKBM (Community Learning Activity Center) from the Ministry of Education, center of business incubator from Ministry of Industry, national entrepreneurship movement from the Ministry of Cooperative and MSMEs, and also in ministry of specific sector such as young generation of agriculture from the Ministry of Agriculture (Hadi, Suardi, and Cahaya, 2015;Mirzanti, Simatupang, and Larso, 2015). However, these programs are not comprehensive yet and require more concrete policy in minister level to generate productive entrepreneurial ecosystem as an effort of internationalization (Hermanto and Suryanto, 2017). Finally, this research has some limitations. First, the finding is limited for MSMEs in Central Java. Most of the MSMEs engaged in this study are those working on handicraft sectors. This kind of MSMEs largely depends on tourism sector to attract consumers. Second, research involving market dynamism aspect are complex ones owing to the connection to strategic management (Donkor et al, 2018). The mechanism joining entrepreneurial cognition, relationship bonding, and network growth might be more complex and can only be reached when the research design employs qualitative method. Next research should use mixed method to reveal the mechanism as well as causal relationship involving mediator and moderator variables in the model, especially for variables in relation to strategic management. Third, this research might have been bias as the data used are those collected from MSMEs practitioners themselves (Ikhsan et al, 2017). Moreover, a cross-sectional design does not facilitate the belief on the causality hypothesized in this research. Thus, the subsequent research needs to use longitudinal design and employ secondary data such as the company"s reports.