The Role of Networking in the Internationalization of Indonesian SMEs

Indonesian SMEs are less able to take advantage of foreign market opportunities than their large counterparts. This study analyses the internationalization of Indonesian SMEs, particularly the differences between exporting and non-exporting SMEs in terms of their network relationships. Primary data was obtained from survey questionnaires in Jawa, Madura and Bali regions, yielding usable responses from 271 exporting SMEs and 226 non-exporting SMEs. Our results suggest that exporters on average have twice as many frequencies of interaction with various external actors than those of their non-exporting counterparts. Exporting and non-exporting SMEs also differ in the way they interact and maintain relationships with external actors. The exporting SMEs utilize various types of interactions including regular and irregular, as well as formal and informal ones, with various external actors in the network. In contrast, non-exporting SMEs are more dependent on personal relations with key persons in various governmental and private institutions. The policy and managerial implications of the findings are discussed.


INTRODUCTION
Benefits of trade openness are not being distributed equally among countries and enterprises. Despite the growing importance of developing countries in the context of global world trade, the 34 OECD member states -representing developed countries -still accounted for 56-62% of world merchandise export value during (ITC, 2019. At the business level, large enterprises are better positioned to capitalize on trade opportunities as compared to their smaller counterparts. For example, during the mid-2000s, small and medium-sized enterprises (SMEs) in the US, Switzerland, the Netherlands, United Kingdom, China and Japan only contributed 30-38% of their respective national exports (Hammer and Stamps, 2010). SME's inability to exploit any such gains from international trade, particularly amidst the context of rapid growth of global trade, indicates that there exist unequal barriers facing enterprises that are contingent upon the firm's scales. Specifically, SMEs face greater impediments and different challenges to internationalize than their larger counterparts, as they have particular characteristics including the smallness and limited resources that may constrain their international business activities (Kahiya and Dean, 2016, Laghzaoui, 2007, Ruzzier et al., 2006, Paul et al., 2017. Owing to the complexity of export activities and SMEs' limited internal resources, most SMEs seek external assistance to deal with various export barriers that they are facing. A substantial number of SMEs are involved in extensive networks orchestrated by business associations, cooperatives, leading companies or foreign buyers in certain industries such as fashion accessories (Battaglia et al., 2006, Johnsen, 2007. These networks or supply chains might provide assistance to their members in reaching the global markets (Lim and Kimura, 2010). Furthermore, network relationships can also be built upon private or social-ties, as is the case with both Chinese SMEs (Zhang et al., 2016, Zhou et al., 2007 and British SMEs (Rodrigues and Child, 2012).
SMEs' meagre export contributions are even more prevalent in developing countries. For example, in ASEAN member states on average SMEs only accounted for 23% of total exports (Wignaraja, 2012). 1 In Indonesia, SMEs only accounted for a minuscule share of 9.3% of total non-oil and gas exports despite being a major source of GDP growth and job creation (Wignaraja, 2012). 2 Despite SMEs' steady rise in total annual export value, their share in Indonesia's non-oil and gas exports shrank continually from around 18.5% in 2005-07 to 16.9% in 2008-10, around 15.4% between 2011 and 2013, and further (2018) and Indonesia-Australia (2019). Indonesia also has ongoing negotiations with several other regional and bilateral FTAs. 1996, Rodrigues and Child, 2012, Shamsuddoha et al., 2009, Wilkinson and Brouthers, 2006.
Intensive research on firm internationalization only began in the late 1950s, as an attempt to explain multinational enterprises' international activities (Buckley, 2011, Dunning, 2006. During this early period there were several seminal theories of firm internationalization including internalization/ transaction cost, eclectic paradigm, and monopolistic advantage theories (Ruzzier et al., 2006). However, these theories are insufficient to explain international activities of smaller enterprises, namely the SMEs. Such differences lie in the fact that SMEs are much less likely to acquire ownership of their foreign counterparts, as they have no superiority over local firms in corresponding foreign markets. In addition, they also lack the capability to pursue vertical integration to further internalize economic activities abroad which is argued in the eclectic paradigm, monopolistic advantage and internalization theories (Hollenstein, 2005, Onkelinx andSleuwaegen, 2008).
Such characteristics contribute to a more dynamic and less deterministic internationalization process among the SMEs. The inaptitude of conventional theories in explaining the SMEs internationalization propel the development of second stream of research on firm internationalization which focuses on the international activities of smaller firms. The more contemporary theories developed within this second stream include the stage model, network model, international entrepreneurship approach and resource-based view. Johanson and Vahlne (1977), Johanson and Vahlne (1990) initially developed a firm internationalization model that would later be known as the Uppsala Model. Within the model, it is argued that firms would take rather small incremental, gradual, and sequential steps in intensifying their engagement in international activities. Firms would begin international activities from foreign markets which have less psychic distance, or perceived as having much similarities to the home market. Generally, firms usually go through three broad stages/phases in their exporting activities: the pre-engagement/preexport stage; the initial/early export stage; and the advanced export stage -in which firms undergo regular exporting activities, coupled with extensive international experience (Leonidou and Katsikeas, 1996).
In contrast with the Stage Theory that highlights the stages of internationalization, Johanson and Mattsson (1988) instead emphasizes the role of network relationships in firms' internationalization process. A network can be defined as "sets of two or more connected exchange relationships" (Easton and Axelsson, 1992). Networks in the international markets could be perceived as systems of cross-border industrial and social relationships among suppliers, customers, competitors, family, and friends (Coviello and Munro, 1997). Specifically, network relationships are particular mutual contracts in which firms establish and assign positions among themselves, whereas such roles might include suppliers, customers, distributors, the industry, and public and regulatory agencies as well as other market actors.
The availability of such relationships might stimulate and facilitate firms to venture abroad. Additional benefits that can be derived from networking include market knowledge, with which firms can gradually expand their activities beyond current scope and across national borders (Johanson and Mattsson, 1988). For the SMEs, network relationships could serve as a fitting strategy and feasible path towards internationalization, as connections in the network will further minimize their need for knowledge development and for adjustment in international markets. Furthermore, the established connections will also help them exploit prevailing business networks within a foreign market (Abdullah andZain, 2011, Ruzzier et al., 2006).
However, some researchers have argued that more and more firms -including the small ones -are having their international orientations developed since their establishment. Such phenomenon enables them to become more internationalized in relatively short-time (Chetty and Campbell-Hunt, 2004). As an attempt to better explain this occurrence of rapid firm internationalization, Oviatt and McDougall (1994) proposed the concept of international new ventures (INVs) or the born global firms. The INV refers to a business organization that derives its competitive advantage from its early-established multinational sale of products and resources.
Finally, the Resource-Based View (RBV) conjectures that a firm's competitive advantage mainly originates from its valuable tangible and intangible resources (Wernerfelt, 1984). In the case of internationalization, it is imperative that a firm carefully considers the question of whether its resources can be a source of competitive advantage, that is, if that firm aspires to enter international markets and to export regularly. More specifically, the firm should assess whether its resources are valuable and unique, in the sense that it remains difficult to be imitated or substituted (Barney, 1991). RBV can be applied to further analyze the international activities of smaller firms, because it helps to gauge the firm's likelihood of success as it attempts to expand in foreign markets. Such analysis can be done without having to follow the internationalization phases which are suggested by the stage theory (Peng, 2001, Sari, 2011. To date, there has not been any single all-encompassing firm internationalization theory that provides an infallible explanation to SME's internationalization, nor is there any such generic theory. This is due to the fact that the process of firms' internationalization and its determinants vary empirically across countries and industries (Onkelinx andSleuwaegen, 2008, Thai, 2008). In other words, such process and its corresponding determinants might potentially be specific to the idiosyncratic nature of each country and industry.
Furthermore, in explaining firms' internationalization process, more recent approaches have been developed, an example of which is the network approach. It is first developed by Johanson and Mattsson (1988) after observing the growing adoption of networking sources by firms to facilitate their internationalization process. Within the context of internationalization, a network can be defined as the firm management team's connection to the firm's customers, suppliers, distributors, competitors, family, friends, bankers, regulatory and public agencies as well as private support agencies, which further enable the firm to expand its business activities beyond domestic markets to overseas counterparts (Coviello and Munro, 1995, Coviello and Munro, 1997, Zain and Ng, 2006. Networking is of paramount importance for SME internationalization, as relationships within the network acts to provide market information, minimize the need for adjustment in a foreign environment, and provide access to established network positions in foreign target markets (Chang and Webster, 2019, Abdullah and Zain, 2011, Ruzzier et al., 2006. Zain and Ng (2006) suggest that network relationships provide supportive roles in the SMEs' internationalization process. Such roles include: 1) prompting and driving SMEs' intention to internationalize; 2) influencing SMEs' selection of target market; 3) influencing SMEs' preferred mode of entry; 4) providing access to broader relationships and established channels within foreign markets; 5) providing access to knowledge of designated market; 6) providing initial credibility within new markets; 7) minimizing the costs and risks associated with business activities overseas; 8) exerting influence on pace and patterns of the internationalization.
However, closer tie with a particular international network can also produce adverse effects in that it might possibly constrain the SMEs' future scope and market expansion opportunity. In some cases, complacent firmsthose that already have close ties with other network members and are more dependent on those members -tend to be less motivated to seek alternative international market opportunities Ng, 2006, Coviello andMunro, 1995). As such, the benefits of market access and international reputation remain concentrated to large firms within the network.
From a theoretical point of view, there are at least three reasons of why network relationships play a crucial role in a firm's internationalization process. First, the aforementioned network model suits the SMEs' internationalization analysis because network relationships may potentially help SMEs to overcome various limitations on internal resources that inhibit them to expand beyond domestic markets. Second, as opposed to the stage theory, which conjectures that internationalization only begins from nearby markets (Ojala, 2009, Zain andNg, 2006), network relationships enable SMEs to begin such process from either close or distant markets. Third, such assistance from network relationships may accelerate the internationalization process of SMEs, contrasting it with the stage theory, which implies the firms' incremental yet gradual progression towards overseas operations (Ojala, 2009).
However, the extant literature differs in classifying the networking sources for entering foreign markets. One commonly used approach is to observe the types of relationships that exist in the international network and subsequently divide such network relationships into formal, informal and intermediary relationships (Ojala, 2009). Formal relationships refer to business in the Internationalization of Indonesian SMEs activities between two or more actors in the international network, while on the other hand, informal relationships refer to personal relationship with relatives, friends and associates. Intermediary relationships can be characterized as having the presence of a third party facilitating the establishment of network relationships between the firm and its foreign buyers or distributors. Other frequently used approach is to look at the networking sources that can be further divided into institutions (including all government agencies), business associates (suppliers, global supply chains, other firms) and personal relations (friends, relatives, colleagues) (Senik et al., 2011).
The existing literature offers a wide variety of views on the manners in which SMEs interact within the networks. Ojala (2009) suggests that SMEs' relationships with actors in the network can develop in either active or passive ways. Active networking is essentially characterized by SMEs' initiatives to build network relationships while passive networking occurs where a network relationship is kick-started by initiatives from the buyers' side. Firms with limited network relationships can assume an active role in building new connections to further facilitate their overseas market expansion. In contrast, network relationships can still occur in spite of passivity on the SMEs' side. Generally, such type of relationships is a by-product of initiatives taken by other actors including customers, importers, distributors or intermediaries. Alternatively, Senik et al. (2011) proposed a network linkage model, in which all networking sources (institutions, business associates and personal relations) work cohesively as a system. The viability of such network linkage system is further safeguarded by the enforcing the facilitation, coordination and monitoring functions of all networking sources. These functions can be performed by a single public body or independent body that connects the potential firms with myriads of public agencies, NGOs, industries, businesses and other actors within the network.
Empirical studies on the role of network relationships in firm internationalization have yielded mixed results. Variations in results are evident across countries, industries, time periods (in which the studies were conducted), forms of internationalization and firm size. The inconclusiveness of network relationships conceptualization, coupled with the problem of fragmented empirical findings further suggest that more research are needed to shed light on this issue.
In the context of Indonesia, to date there has been an extensive body of literature on Indonesian SMEs, but only few sheds light on small firm internationalization. Of those few studies, even less researchess have addressed the internationalization networking issue. For example, Revindo and Gan (2016) investigated how the presence of network sources may stimulate Indonesian SMEs to initiate export activities. It has also been argued that the relationships with external actors determine SMEs' engagement in export market (Revindo and Gan, 2017) and SMEs' export intensity (Revindo and Gan, 2018). However, to the extent of our knowledge there has been no study that specifically investigates types of network sources that the SMEs utilize and the types of interactions engaged by SMEs with external actors in their effort to internationalize. This study aims to fill the gap by investigating how Indonesian SMEs develop network relationships to participate in export markets.
The remainder of the paper is organized as follows. Section 2 presents the method including study area, the data collection procedure and the data analysis method. Section 3 discusses the empirical results. Finally, Section 4 concludes with a summary of the main research findings and the research implications.

METHOD
This study narrows its focus on smallsized and medium-sized enterprises hence excludes their micro-sized and large-sized counterparts 5 . Hereafter, SMEs are defined by the number of employees, which ranges from 5 to 99 -in accordance with formal definition from BPS-Statistics Indonesia. It is estimated that 60% of the currently existing SMEs in the country remains concentrated within 3 islands; namely, Jawa, Madura and Bali (Sabila, 2014).
To construct the sample frame, we merged four different databases into one comprehensive list of SMEs, from which the samples were then picked. The first three databases were published by the Ministry of Cooperatives and SMEs including: (1)  First, the micro enterprises database is unavailable in Indonesia as they are mostly in the form of individual businesses or home industries. Second, micro enterprises are less likely to engage in international business (Pendergast et al., 2008). 6 Online promotion at the website of the Ministry of Cooperatives and SMEs: http://www.indonesian-products.biz. 7 The catalogue provides SMEs' contacts and products description in four languages (English, Arabic, Japanese and Indonesian). The catalogue is published annually as part of the ministry's promotion program. 8 The directory books listed all SMEs that participated in international trade shows organised by the Ministry of Cooperatives and SMEs' during 2005-2009. the Indonesian 2006 Economic Census provided by BPS-Statistics Indonesia. 9 The survey was administered in 2014, during which 971 SMEs were contacted and 497 of which responded (around 51% response rate).
In the survey, the SMEs' owners/managers were asked to indicate whether they had received any external assistance to help them overcome various export barriers -including financial, informational, marketing, distribution, human resources, product, procedure and business environment barriers. The respondents were then asked to identify the sources/providers of such assistance, if any, which include: 1) central government agencies; 2) local government agencies; 3) business association/ chambers; 4) private companies/ state owned enterprises (SOEs); 5) universities/research institutes; 6) businesspartners/ associates; 7) Indonesian emigrant communities overseas; and 8) family/relatives (Battaglia et al., 2006, Senik et al., 2011. The respondents were also asked to indicate the helpfulness of assistance received in the Internationalization of Indonesian SMEs using a three point Likert-Scale (1 = not helpful, 2 = helpful, 3 = very helpful).
The results from the survey are then analyzed as follows. First, we identify the type of networking sources that provide the most assistance (indicated by the provision frequency of such assistances, as reported by the SMEs) and those that provide the most helpful assistance (indicated by highest average scores of helpfulness). Second, we also identify the various types of assistance that are being provided the most by the various networking sources (indicated by count of assistance received by the SMEs) and those that are perceived as being the most useful (scoring highest average scores of helpfulness).
The survey also inquires the forms of relationships that the SMEs established and maintains with the aforementioned eight types of networking sources. Furthermore, such forms of relationships being considered include seven formal relations and two informal relations (Senik et al., 2011). Examples of formal relationships being maintained by the SMEs include: being regular participants in various assistance programs, being irregular participants in all assistance program, making regular contact through formal/official discussions/seminars, making irregular contact through formal/official discussions/ seminars, being a member of a forum set up by agencies /associations/institutes, and involvement in strategic partnership(s). Examples of informal relationships include personal relationships with key persons in the agencies/institutions and making indirect contact with agencies/institutions through another party. The independence test (Chisquare test) is employed to draw the association, between SMEs' export status (exporting and non-exporting SMEs) and the form of relationships that they maintain with various networking sources.

RESULTS AND DISCUSSION
SMEs gain the information needed to identify export opportunities from various sources. In the survey, respondents who identify themselves as being exporters or aspiringexporters were asked to identify sources of information with which they obtain information regarding export opportunities in foreign target markets. Table 1 shows that both exporters group and their aspiring counterparts use ten sources to obtain such information. However, the Chi-square test performed on comparing both groups indicates that these groups differ in the extent to which they use the individual sources of information.
Business partners/associates and potential buyers are two most utilized sources of information for both SMEs groups. However, the exporters report that business associations and web resources are the third and fourth most important sources of information, while such order are reversed for the aspiring exporters. Further, central government agencies are ranked by the exporters as the fifth most important source of information while for the aspiringexporters it is the media (newspapers, televisions and internet) and family/relatives that rank as the fifth most important source of information. This suggests that government agencies still have further rooms to improve its role in providing informational assistance for SMEs at pre-export stage. Overall, these findings also indicate that exporting SMEs appear to have better network relationships with business associations and central government agencies than its aspiringexporter counterparts. Note: (***) represents a 1% significance level The counts exceed the number of respondents because each respondent may identify more than one source of information Source: Author's calculation based on survey data Table 2 shows the types of relationships maintained by SMEs with various sources of networking. It shows that different networking sources correspondingly entail different types of relationships. The type of relationships being maintained between SMEs and central government agencies mainly occurs in the form of participation in programs designed specifically to support SMEs. Around 28.6% of respondents interacting with central government agencies stated that they have been regular participants of such programs. Other notable forms of interactions include having personal relations with key persons within the government agency (16.4%) and being irregular participants of the same programs (16%).
The same phenomenon can be observed in the context of Regional Government Agencies, whereas most of the interactions made between the SMEs and these government agencies also occur in the form of participation in various SMEs supporting programs. However, the data also shows that percentage of regular participation in the regional government's programs is significantly lower (by almost 7 percentage points) than its central counterparts. Such observation suggests further room for the Regional Government to increase its engagement in the process of supporting the SMEs.
SMEs are most engaged with business association/chambers in the form of being members of forums set up by these institutions. For this type of interaction, business associations have proven to be prominent when compared to other networking sources (highest percentage of 19.7%). Other notable forms of interactions also include having personal relations with key persons within the association (16%) and being regular participants in various SMEs supporting programs (13.7%).
Interactions being made between the SMEs and universities/research institutes mainly occur in the form of being irregular participants in all supporting programs for SMEs (21.5%), indicating the presence of spillover effect from technological developments being made within the academic institutions. However, such interactions are not regularly attended -suggesting further role for improving this type of interaction in order to magnify the spillover effect. Other noteworthy types of relationships also include having indirect contact through other party (21.5%) and having personal relation with key persons. In relation to private companies/ SOEs, the types of relationships being maintained with the SMEs are mostly in forms of having personal relation with key persons (18.8%). Other noteworthy types of interactions include being both regular and irregular participants in all SMEs supporting programs (16.8% and 14.7%, respectively), and the development of strategic partnerships (11.7%). Moreover, results from horizontal comparison -indicating the relative importance of a networking source for a given type of relationship -demonstrate the fact that these companies still serve minor roles in providing network assistance for the SMEs, as their percentages are significantly lower than that of other networking sources.

Revindo, Indrawati & Hambali, The Role of Networking in the Internationalization of Indonesian SMEs
Types of relationships between the SMEs and Business Partners/Associates are mostly manifested in the form of strategic partnerships (28.5%). Other types of relationships that are worth noting include informal relationship -in the form of indirect contact through other party (25.2%) and personal relation with key persons (22.9%), and joint projects between the two business entities (22.9%). It is interesting to note that the business partners/associates have proven to exert significant role in providing network assistance to the SMES. For example, they are most prominent in establishing strategic partnerships and joint projects with the SMEs (45.9% and 46.7% compared to other networking sources). This might be attributed to their smaller scale as compared to the private companies/SOEs, which provides them with a competitive edge in coordinating with the grassroots companies.
The same phenomenon can be observed for the family/relatives source of networking. A horizontal comparison also demonstrates the fact that the SMEs also counted on their families/relatives to provide them strategic partnerships (23.7% compared to other networking sources). In addition, this networking source is also counted on by the SMEs to establish joint projects (17.6%).
Meanwhile, the data also shows that emigrant communities have played an astonishingly minor role despite the importance of their role in bridging the gap between international and domestic markets. Table 3 displays the differences between exporting and non-exporting SMEs in terms of frequencies of contact/interaction made with these eight types of networking sources. There are observed similarities and differences among exporting SMEs and their nonexporting counterparts, particularly in terms of types of networking sources being accessed. Both of these groups have the most frequent network access to business partners/associates and business associations/chambers and the least frequent interaction with Indonesian overseas diaspora communities.
One of the stark differences between the exporting and non-exporting SME groups lies in the number of interactions being reported, as the former reports 1,144 interactions as compared to 694 by the latter. Table 3 below also infers that exporting SMEs exhibit a more balanced interaction between central and local government agencies (15.30% and 15.12%, respectively), whereas non-exporting SMEs tend to interact more closely with local government agencies as compared to the central counterparts (18.30% and 9.08%, respectively). This may potentially be driven by the fact that local government agencies tend to possess stronger local market orientation in extending their assistance, while the central government agencies may have a more global vision in identifying market opportunities for SMEs (Uchikawa and Keola, 2008). Note: (***) represents a 1% significance level The counts exceed the number of respondents because each respondent may identify more than one source of information Source: Author's calculation based on survey data Note: (***) represents a 1% significance level The counts exceed the number of respondents because each respondent may identify more than one source of information Source: Author's calculation based on survey data

Revindo, Indrawati & Hambali, The Role of Networking in the Internationalization of Indonesian SMEs
We then investigate the differences between exporting and non-exporting SMEs in terms of channels with which these SMEs build and maintain interactions with various networking actors/sources. Table 4 displays the nine types of formal and informal interactions made between the SMEs and their networking sources, ranging from intense and regular interaction (Regular participant in all supporting programs for SMEs) to informal and irregular interaction (Indirect contact through other party). Table 4 shows that exporting and nonexporting SMEs differ in the channels they utilize to maintain relationship with external actors, as reflected by the significant chisquare statistics (at 1% significance level). Exporting SMEs make use of both regular and irregular as well as both formal and informal interactions to maintain existing networks. On the other hand, non-exporting SMEs tend to rely more on personal relations with key persons in governmental/private institutions. In addition, strong tendency to rely on indirect contact through other party is also observed among the non-exporting SMEs. These differences serve as one of the key factors that distinguish the exporting SMEs and their non-exporting counterparts.
We further explore the means with which the exporting SMEs use networking sources to assist them in their export activities. Table 5 summarizes the number of export-related assistance that these SMEs received from their networks for each export task/function. The last two columns in Table  5 show the ranks and the average helpfulness score of each type of assistance received. The latter is measured on a three-point Likert Scale (1 = not helpful, 2 = helpful, 3 = very helpful).
From Table 5, one can derive the conclusion that most of the assistance given to the exporting SMEs comes in the form of informational assistance regarding foreign market export opportunities, which amounts to 636 counts (21% of total assistance counts). Other noteworthy assistance provided by external actors in the network include marketing and promotional activities (299 counts), working capital (273) and improvement in managerial team capabilities (259). On the other hand, export-related activities that have been receiving the least attention include unequal treatment in foreign markets (69 counts of assistance), export guarantees and insurance (71), contracts and dispute settlement with foreign customers (90) and anticipation of foreign market turbulent economic conditions (93).
However, the counts of assistance do not necessarily provide information on the helpfulness of the assistance being provided. The last column in Table 5 shows that assistance on the capital needs of these SMEs are perceived to be the most useful assistance, scoring an average of 2.48. Other assistances strongly perceived to be useful also include information related to export opportunities (with an average score of 2.47) and promotional activities (2.46). On the other hand, assistances that are least perceived to be useful are generally informational assistance with regards to business environment functions, such as unequal treatment in foreign countries (2.27), followed by informational assistance on differences in business practices at the destination market (2.28) and differences pertaining to regulation in foreign countries (2.29). Note: The counts exceed the number of respondents because each respondent may receive assistance in more than one export function. Source: Author's calculation based on survey data

Revindo, Indrawati & Hambali, The Role of Networking in the Internationalization of Indonesian SMEs
We then identify the types of networking sources that most actively provide export-related assistance and the most useful assistance, as perceived by the exporting SMEs. The findings are summarized in Table 6 below.
From the table, we observe that business partners and associates provide the most export-related assistance for the SMEs, amounting to 1,016 counts of assistance for all 16 export functions. Other noteworthy networking sources that provide the most assistance for SMEs include central government agencies (499) and business associations/chambers (410).
In contrast, we also note that universities/research institutes and Indonesian emigrant communities abroad provide the least export-assistance (59 and 98 respectively). This is indicative of the rather weak partnerships between research institutes/universities and SMEs, particularly if one considers the crucial role that these research bodies can play in creating innovations and developing cutting-edge products. Likewise, the relatively lack of support from diaspora communities is also worth mentioning, since their role in further connecting domestic SMEs to foreign markets are also of paramount importance.
The last column in Table 6 shows that the assistance provided by private companies/state-owned enterprises and assistance from family/relatives are perceived by SMEs to be the most helpful. On the other hand, it is worth noting that the assistance provided by central and local government agencies neither ranks on top of the assistance provision counts nor on top of the helpfulness score. There are at least two ways in which government agencies can improve their currently existing provision of export assistance. First, the government may consider increasing assistance provision related to the export functions that still have low assistance counts. In other words, the government may choose to focus on providing assistance in export functions that have not been adequately assisted by other networking sources. In this case, the government may concentrate their efforts in giving informational assistance related to the business environment or barriers in export procedures.
Alternatively, the government can also choose to focus on assisting export functions that still rank low in terms of average helpfulness score. In other words, the government may focus on providing assistance in export functions for which the current types of assistance from various networking sources are still ineffective. In this case, the government may concentrate their effort in addressing the business environment issues.
However, it can also be argued that the government does not always have to intervene in the network relationships activities (Yamin and Ghauri, 2004). In our study context, the government may refrain from being directly involved in every export function (i.e. providing assistance for each export function). Instead, the government may coordinate, facilitate or strengthen the relationship between SMEs and their business partners/associates, private companies/SOEs and business associates/ chambers (Senik et al., 2011). Such example would be facilitating the meetings, partnerships or cooperation between SMEs and their various networking sources. Hence, rather than intervening heavily by designing export assistance programs, the government may instead enhance existing networks in order to ensure that they function efficiently in providing assistance to the exporting SMEs.   Table 7 details the number of export assistance that SMEs received from the networks for various export tasks/functions. With regards to export assistances received from central government agencies, the SMEs stated that these government agencies mostly provide them with information on business opportunities abroad (87 of all assistance provided by the agencies). In addition, there are several other forms of export assistance that have been provided more by the government and less by other networking sources. Such assistances include those related to export procedures and marketing functions. The former aims to provide the SMEs with relevant information regarding export procedures and other paperwork needed in the export process. On the other hand, the latter is more concerned with countering negative images of Indonesian products.
Similar to their central government counterparts, the regional government agencies also mostly provide the SMEs with information related to business opportunities abroad (101 of all assistance provided). Such assistance is also higher in frequency compared to those provided by their central government counterparts, which might be possibly due to the regional government's closer proximity to the SMEs. Moreover, the table also clearly shows that regional government agencies specialize in providing human resources-related assistances (18.1% of all human-resources related assistance).
Being the highest assistance-providing network source, most of the assistances provided by business partners/associates are also related to information on opportunities abroad. However, it is noteworthy that this institution provides the highest number of opportunity-related information as compared to other networking sources. Other assistance mostly provided by the association/chambers are related to product development (101 of all assistance provided) and marketing functions (91 of all assistance provided). In contrast, universities/research institutes provide the least number of assistance to SMEs. If any, such assistances occur in forms of human-resource related or marketing-related assistances.
Private companies/SOEs also play a crucial role in the process of assisting the SMEs, as they mostly channel assistances perceived to be most helpful by the SMEs. Such assistances are related to meeting the financial needs of the SMEs. Of all finance-related assistances, around 27.8% are provided by private companies/SOEs. The prowess of these companies in providing SMEs with capital access are arguably due to their large business scale, and therefore capital base, which further enables them to engage in venture capital practices. Other noteworthy assistances provided by these companies also include information-related (22 of all assistance provided), distribution-related (16), and marketing-related assistances (16).
As with the government entities and business partner, business associates/chambers mostly assists the SMEs by providing information on opportunities abroad (75 of all assistances provided). Other assistances being provided by the associates/chambers include marketing-related (44) and human-resource related assistances (39). Those closest to the SMEs managerial boards, that is, their family/relatives, are counted on to provide finance-related export assistance. Such phenomenon can be attributed to strong sentiment of kinship prevalent within Indonesian grassroots business communities, which further enables the SMEs to receive loans at lower or zero interest rates.
Despite the importance of bridging domestic players with international markets, Indonesian emigrant communities are among the lowest assistance-providing network sources. Its total assistance to SMEs only amounts to 98, as compared to 1,016 assistances provided by business partners/associates. Most assistances provided by these diaspora communities come in form of information related to business opportunities abroad (26 of 98 assistances being provided). However, the number of such assistances can be further magnified to better target foreign markets and thereby increasing the profitability of exporting activities.
The above-elaborated results suggest that exporting SMEs, as with the case for aspiring-exporters, should also seek to maintain and strengthen their relationships with various non-government actors within the network. These exporting SMEs may seek advice or information related to export expansion opportunities from business associations/chambers and business partners/associates. In addition, the current exporters should also build networking with Indonesian emigrant (diaspora) communities to provide them further access to multiple foreign markets. The role of the Indonesian diaspora is still very limited and ineffective in facilitating SME internationalization when compared to other communities, such as guanxi (China), kankei (Japan) and immak (Korea) (Zhou et al., 2007, Zhang et al., 2016. The results further provide several academic implications. Our findings are consistent with the Network Model of internationalization. In line with Johanson and Mattsson (1988), we found that network relationships are of crucial importance for SME internationalization, as indicated by the fact that exporting SMEs did indeed build more intensive interactions and close relationships with various government agencies and non-government actors than did the non-exporting SMEs. Through these external actors, SMEs attain reliable information on export opportunities and foreign market. Such external actors include business associates/partners, foreign buyers and business associations/ chambers. Accordingly, we also observed that SMEs require financial, managerial, technical and promotional assistance from both central government agencies and non-government sources to become exporters, and to further sustain and expand their export activities.

CONCLUSION
Network relationships play a crucial role in facilitating the internationalization of Indonesian SMEs. Our results suggest that exporters have twice as many frequencies of interaction with external actors than those of their non-exporting counterparts. Such external actors include central and local government agencies, business associations/chambers, universities/research institutes, private companies/SOEs, business partners/associates, family/ relatives and Indonesian emigrant communities overseas. Exporters and non-exporters also differ in the types of network sources that they mainly utilize. The exporting SMEs show close interactions with both central and local government agencies, whereas non-exporting SMEs are more inclined to establish close interactions with local government agencies than to central government institutions.
Further, exporting and non-exporting SMEs also differ in the way they interact and maintain relationships with external actors. The exporting SMEs utilize various types of interactions including regular and irregular, as well as formal and informal ones, with various external actors in the network. In contrast, nonexporting SMEs are more dependent on personal relations with key persons in various governmental and private institutions.
External actors in these SMEs' networks provide export assistance that is mostly related with the supply of information regarding foreign market business opportunities, marketing and promotional activities. Moreover, we find that assistance associated with unequal treatment in foreign markets, export guarantees and insurance, and contracts and dispute settlement (with foreign customers) remains small in amount. However, we find that the type of assistance perceived as the most helpful by these enterprises is assistance in terms of working capital. Conversely, these enterprises perceive assistance associated with business environment functions (including unequal treatment in foreign countries) as being the least helpful assistance.
Our results also suggest that the networking sources which provide the most export assistance to SMEs are business partners/associates, central government agencies and business associations/ chambers. On the other hand, the sources that are perceived to provide the least export assistance are universities/research institutes and Indonesian diaspora communities overseas. It is worth mentioning that the frequency of assistance being provided does not necessarily reflect the helpfulness of such assistance. As such, our study suggests that SMEs perceive assistance provided by private companies, state-owned enterprises and family/relatives as being the most beneficial for them.
Government agencies also play a paramount role in facilitating the internationalization of SMEs. We observed that exporting SMEs receive more assistance from the central government as compared to their non-exporting counterparts in all types of assistance, including international trade fairs, publication of SME Catalogues, managerial training, technical training, and export financing, insurance and guarantees.
The findings of the study have several important policy implications. The government should be more informed of the functions and role of non-government actors within the internationalization network. Such non-government actors include business associations/chambers, research institute/ universities, finance/microfinance institutions and other non-government organizations. Moreover, government intervention should incorporate the network relationships between SMEs and other actors that are already at work.
Thus, the government should define the appropriate level of intervention for each policy area. Direct intervention (export assistance provision) might be suitable for export tasks that have not yet been adequately provided by nongovernment actors within the network. An example might be the intensification of efforts by the government to address issues of business environment barriers in foreign markets, such as SMEs' insufficient knowledge of target markets' regulations, procedures and business practices, and the negative image of Indonesia or Indonesian products. Such types of barriers are beyond non-government actors' capability to deal with, and should be within the government's scope of action.
On the other hand, in the case where nongovernment actors have provided adequate assistance to these exporting enterprises, indirect intervention may then be preferred over direct intervention. As such, it is not imperative, for example, that the government establishes a new form of financial institution or any new schemes to support SMEs export. Rather, the government may instead opt to strengthen the operation of currently existing non-government actors, such as the Indonesia Eximbank, to increase its service outreach and to raise SMEs' awareness of the bank's services.
Example of other alternatives include the assignment of one public body by the government to facilitate, connect, coordinate and monitor the myriads of private and public agencies that possess the same area of interest or assistance (Senik et al., 2011). For example, the government may assign the Directorate General