Journal of International Business Studies (2012) 43, 458–476

& 2012 Academy of International Business All rights reserved 0047-2506

www.jibs.net

Asset specificity and foreign market entry

mode choice of small and medium-sized
enterprises: The moderating influence of
knowledge safeguards and institutional
safeguards

Birger Maekelburger1, Abstract

Christian Schwens2 and According to transaction cost economics (TCE) reasoning, firms choose equity
(as opposed to non-equity) foreign market entry modes to safeguard specific
Ruediger Kabst1 assets. The present paper contextualizes the well-researched relationship
between asset specificity and foreign market entry mode choice by introducing
1
Department of Human Resource Management, knowledge safeguards (international experience, host-country networks, and
Small and Medium-Sized Enterprises, and imitation) and institutional safeguards (property rights protection and cultural
Entrepreneurship, University of Giessen, Giessen,
proximity) as alternative mechanisms for securing a firm’s specific assets.
Germany; 2Faculty of Business Administration
and Economics, Heinrich-Heine University Testing our hypotheses on a sample of 206 small and medium-sized enter-
Düsseldorf, Düsseldorf, Germany prises, we find that knowledge safeguards and institutional safeguards weaken
the effect of asset specificity on the choice of equity foreign market entry
Correspondence: modes. Contextualizing the relationship between asset specificity and foreign
B Maekelburger, Faculty of Economics and market entry mode choice helps to enhance our understanding of the scope
Business Administration, Department of conditions of TCE-based entry mode studies and beyond.
Human Resource Management, Small- and Journal of International Business Studies (2012) 43, 458–476.
Medium-Sized Enterprises, and
doi:10.1057/jibs.2012.12
Entrepreneurship, University of Giessen,
Licher Str. 62, 35394 Giessen, Germany.
Tel: þ 49 641 992 2106; Keywords: foreign market entry mode choice; small and medium-sized enterprises
Fax: þ 49 641 992 2109; (SMEs); asset specificity; knowledge safeguards; institutional safeguards
email: birger.maekelburger@wirtschaft
.uni-giessen.de
INTRODUCTION
Control is the ability to influence systems, methods, and decisions,
and determines the level of risk and return (Anderson & Gatignon,
1986). Entry mode choice refers to the degree of control a firm
desires when venturing into a foreign market (Pan & Tse, 2000).
The entry mode literature typically discusses control from the
standpoint of structural equity (Gatignon & Anderson, 1988). Pan
and Tse (2000) explain foreign market entry mode choice as a
hierarchical decision-making process in which the choice between
equity entry modes (equity joint venture and wholly owned
subsidiary) and non-equity entry modes (export and contractual
Received: 18 May 2010
Revised: 14 July 2011
agreement) is at the top level. The choice is of paramount
Accepted: 20 January 2012 importance to small and medium-sized enterprises (SMEs). Because
Online publication date: 3 May 2012 of their severe resource constraints, SMEs have difficulty choosing
 Asset specificity and SME entry mode Birger Maekelburger et al
459

equity entry modes, even if they are efficient ability to reduce internal uncertainty by using
(Erramilli & D’Souza, 1993). international experience, host-country networks,
Williamson’s (1985, 1998) transaction cost eco- and the imitation of best-practice firms in the
nomics (TCE) framework is the economic theory host country. SMEs can benefit from institutional
of choice for many entry mode studies (Brouthers & safeguards by choosing a host-country market with
Hennart, 2007). The underlying discriminating strong property rights protection and high cultural
alignment hypothesis suggests that transactions, proximity, exposing the firm to low levels of
which differ in their attributes, are aligned with external uncertainty. The TCE framework suggests
governance modes, which differ in their associated that idiosyncratic investments require protection
costs and competences, maximizing efficiency through internalization when asset specificity is
(Henisz & Williamson, 1999). The default hypo- coupled with uncertainty (Geyskens et al., 2006).
thesis suggests that a low level of ownership is Our central argument is that SMEs can rely
favorable until it is proven otherwise (Anderson & on alternative safeguard mechanisms for foreign
Gatignon, 1986). Of the three central TCE attri- market entry, which make expensive equity entry
butes (asset specificity, uncertainty, and frequency), modes less necessary.
asset specificity maintained its significance as the We offer two contributions to the extant litera-
key determinant of entry mode choice (Henisz, ture. First, we contextualize the central relationship
2000b). Asset specificity refers to a transaction between asset specificity and foreign market entry
partner’s specific investments that lose value in mode choice by investigating the effect of varying
alternative use (Williamson, 1991). Whereas some levels of knowledge safeguards and institutional
studies find a positive relationship between asset safeguards. We provide theoretical explanations
specificity and the level of equity in entry mode for how these alternative mechanisms can secure
choice (e.g., Brouthers & Brouthers, 2003; Erramilli a firm’s specific assets in the context of foreign
& Rao, 1993; Gatignon & Anderson, 1988), others market entry mode choices. Prior research has pro-
cannot or not consistently replicate these findings vided valuable knowledge regarding the safe-
(e.g., Delios & Beamish, 1999; Hennart, 1991; guarding effect of internalization, leaving room
Taylor, Zou, & Osland, 1998). for the discussion of alternative safeguards that
Contextualizing the relationship between asset will further our understanding of the theory.
specificity and foreign market entry mode choice Second, we study the interaction effects of asset
helps to enhance our understanding of the scope specificity with knowledge safeguards and insti-
conditions of this relationship. Contextualization tutional safeguards in the context of SMEs. Thus
is imperative when studying SMEs. In addition we add to the literature on the foreign market
to resource constraints, SMEs are subject to two entry mode choice of SMEs (Erramilli & D’Souza,
further liabilities of smallness. First, they tend to 1993, 1995; Li & Qian, 2008). The knowledge
lack internationalization knowledge (Lu & Beamish, safeguards and institutional safeguards developed
2001). A lack of experience with or knowledge of in the present paper are specifically tailored to the
foreign markets makes it difficult for SMEs to verify three liabilities of smallness: resource constraints,
a business partner’s performance ex post (Zhao, Luo, lack of internationalization knowledge, and sensi-
& Taewon, 2004). This difficulty leads in turn to tivity to external influences. However, although
internal uncertainty, ranging from honest disagree- these liabilities are typical for SMEs, they are not
ment to opportunism (Hesterly & Zenger, 1993). exclusive to small firms. Hence we envisage the
Second, SMEs are easily influenced by their envir- alternative safeguarding rationale discussed in the
onment (Cheng & Yu, 2008), and react sensitively present paper to be applicable to certain larger
to institutional challenges (Schwens, Eiche, & firms, and believe that our findings may inform
Kabst, 2011), which makes it difficult to predict future TCE-based studies addressing foreign market
future events. The inability to specify the envi- entry mode choice and beyond.
ronment ex ante leads to external uncertainty
(Geyskens, Steenkamp, & Kumar, 2006). BACKGROUND LITERATURE
The aim of the present paper is to contextualize
the asset specificity and foreign market entry Asset Specificity and Foreign Market Entry Mode
mode choice relationship in the context of SMEs Choice
by studying two sets of alternative safeguards. Based on TCE arguing, mode choice is a critical
Knowledge safeguards describe an SME’s inherent decision of governance (Zhao et al., 2004). The

Journal of International Business Studies

 Asset specificity and SME entry mode Birger Maekelburger et al
460

theory distinguishes governance modes along acknowledges social and institutional contexts,
a continuum that extends from markets to hier- and extended TCE to better examine the gov-
archies (Henisz & Williamson, 1999). TCE cha- ernance of knowledge. Mayer and Salomon (2006)
racterizes governance modes according to their addressed the lack of contextualization by combin-
different levels of resource commitment, which ing TCE with resource-based arguments to explain
in turn define the associated levels of control and governance, whereas Oxley (1999) extended TCE
risk (Pan & Tse, 2000). TCE rests on two key toward institutional issues. Although scholars have
assumptions: bounded rationality and opportu- attempted to contextualize TCE, the entry mode
nism (Williamson, 1985). Because of bounded literature is still limited with regard to a contextua-
rationality, contracts are unavoidably incomplete, lization of the relationship between asset specificity
and because of opportunism, they are not self- and foreign market entry mode choice (Brouthers
enforcing (Henisz & Williamson, 1999). & Hennart, 2007), a gap that we strive to close. By
The main action of TCE is the discriminating linking TCE with both knowledge safeguards and
alignment hypothesis, which suggests that trans- institutional safeguards, we shed further light on
actions, characterized by specific attributes, are the alignment between asset specificity and foreign
aligned with governance modes, which themselves market entry mode choice.
are characterized by certain costs and competen-
ces (Henisz & Williamson, 1999). This alignment SMEs and the Liabilities of Smallness
allows firms to coordinate incentives and efforts to Contextualization is of paramount importance
achieve efficiency gains (Mesquita & Brush, 2008). for SMEs. Although some scholars have shown
The a priori TCE assumption suggests that market TCE to work well in predicting foreign market entry
governance is generally preferable, because internal mode choice for both large multinational enter-
transactions are often insulated from competitive prises (MNEs) and SMEs (e.g., Brouthers & Nakos,
pressures, and hence are vulnerable to bureaucratic 2004), others have found that contextual factors
phenomena (Geyskens et al., 2006). However, the are important in distinguishing SME entry mode
three attributes characterizing transactions – asset choice from that of large MNEs. Erramilli and
specificity, uncertainty, and frequency (Williamson, D’Souza (1993) found that the differences between
1985) – create transaction costs, and may make smaller and larger firms are negligible when capital
internalization more efficient and hence more intensity is low, but become substantial when
attractive (Geyskens et al., 2006). capital intensity increases, and that the effect of
The term “asset specificity” is used to describe external uncertainty on entry mode choice is
investments that are specific to a transaction, and moderated by firm size. Other empirical evidence
which lose value in alternative use (Williamson, showed that smaller firms shy away from foreign
1991). The idiosyncratic nature of these investments direct investment (FDI) in situations characterized
makes them vulnerable to opportunism, and safe- by high degrees of country risk, whereas larger firms
guarding becomes an important issue (Geyskens prefer higher levels of FDI in the same situations
et al., 2006). Because market competition cannot (Erramilli & D’Souza, 1995). Finally, Li and Qian
control opportunism, firms implement gover- (2008) found that in situations involving high
nance mechanisms to protect their investments levels of violent market dynamism, the entry mode
from losses, haggling, and negotiation inefficiency choice of smaller firms was different from that of
(Mesquita & Brush, 2008). A common safeguarding large firms.
strategy is internalization (Teece, 1986). Although These results are not surprising, given the liabi-
some entry mode studies find support for a positive lities of smallness. First, SMEs lack important
relationship between asset specificity and equity resources (e.g., Buckley, 1989; Calof, 1994; Cheng
foreign market entry modes (e.g., Brouthers & & Yu, 2008). When resources are low, so are error
Brouthers, 2003; Erramilli & Rao, 1993; Gatignon margins. The relative proportion of resources
& Anderson, 1988), others cannot or not con- committed to a single foreign market is likely to
sistently confirm the existence of this association be higher for smaller firms than for larger firms,
(e.g., Delios & Beamish, 1999; Hennart, 1991; making failure more costly (Buckley, 1989). In fact,
Taylor et al., 1998). risking specific and non-redeployable investments
From early on, TCE was criticized for its lack of may jeopardize future growth or even the survival
contextualization (Granovetter, 1985). Mayer (2006) of the SME (Sapienza, Autio, George, & Zahra,
emphasized the limited extent to which TCE 2006). According to TCE-based reasoning, firms

Journal of International Business Studies

 Asset specificity and SME entry mode Birger Maekelburger et al
461

choose the entry mode that is best suited to International Host-country

Imitation
safeguarding specific investments against potential experience networks

opportunistic behavior. However, resource con-

H1a- H1b- H1c- Foreign market
straints severely restrict SMEs in their choice of Asset entry mode
specificity
equity entry modes, even when internalization is H2a- H2b-
(equity vs
non-equity)
rational (Erramilli & Rao, 1993).
Second, SMEs generally lack knowledge and Property rights
protection
Cultural
proximity
experience related to internationalization (Lu &
Beamish, 2001). SMEs with inexperienced mana-
Figure 1 Research model.
gers, for instance, inevitably lack the legal,
social, and political discernment to operate abroad
(Buckley, 1989). This problem is aggravated by
the first liability; smaller firms typically do not
have sufficient financial resources to hire and train in the host country (Henisz & Delios, 2001) and
international personnel (Calof, 1994). TCE suggests cultural proximity between the home and host
that internal uncertainties arise when companies country (Slangen & van Tulder, 2009) are appro-
are unable to accurately assess a business partner’s priate institutional safeguards against external
performance (Anderson & Gatignon, 1986). This uncertainty.
issue often manifests itself in a lack of relevant In the following, we develop the rationale for
knowledge and experience (Zhao et al., 2004). the influence of knowledge safeguards and insti-
Third, SMEs are highly sensitive to external tutional safeguards on the relationship between
influences (Cheng & Yu, 2008). They are generally asset specificity and SME foreign market entry
vulnerable to technological, political, institutional, mode choice, as illustrated in Figure 1.
and market changes (Buckley, 1989). Hence, the
challenges arising from the host-country environ-
ment strongly influence the foreign market entry
HYPOTHESES
mode choice in SMEs (Cheng & Yu, 2008). TCE Knowledge Safeguards
suggests that external uncertainty prevails when
companies cannot predict future events. Henisz International experience
and Williamson (1999) highlight the lack of pro- We define international experience as the degree to
perty rights protection as an important formal which an SME and its management have been
source of external uncertainty, and Brouthers involved in cross-border business activities (Burgel
and Hennart (2007) suggest that cultural distance & Murray, 2000). Lack of knowledge and experience
is the most common informal source of external is a source of internal uncertainty (Zhao et al.,
uncertainty. 2004). The underlying rationale is that firms that
Limited resources can restrict SMEs in choosing are new to a foreign market typically have not
equity entry modes, even when internalization developed methods for gathering and analyzing
is rational (Erramilli & Rao, 1993). Therefore we relevant information, and have not implemented a
introduce two sets of alternative safeguards that heuristic for making investments (Henisz & Delios,
allow SMEs to protect their specific assets. First, 2001). This may lead to opportunistic behavior by
knowledge safeguards describe an SME’s inherent business partners. Experience provides information
ability to reduce internal uncertainty. According about potential transaction partners in the host
to previous research, international experience country, which can reduce the “expected variance
(Delios & Beamish, 1999; Henisz & Delios, 2001), in the opportunistic behavior” (Delios & Henisz,
host-country networks (Provan, 1993), and the 2000: 309) that jeopardizes the SME’s specific
imitation of best-practice firms in the host country assets. Thus international experience enhances
(Forsgren, 2002) are knowledge mechanisms that understanding, competence, and confidence, and
SMEs can use to reduce internal uncertainty. improves projections regarding risks and returns
Second, SMEs can benefit from institutional safe- (Gatignon & Anderson, 1988). Firms can utilize
guards by choosing a host-country with limited experience from previous foreign endeavors to
formal and informal institutional challenges, and safeguard against potential opportunistic beha-
hence low external uncertainty. According to pre- vior by their partners, making high control
vious research, sufficient property rights protection modes less necessary, despite their specific assets

Journal of International Business Studies

 Asset specificity and SME entry mode Birger Maekelburger et al
462

(Henisz & Delios, 2002).1 This argument leads us country perceived as best practice (Henisz & Delios,
to develop the following hypothesis: 2001). Imitation makes internalization of specific
assets less necessary, because it renders the SME’s
Hypothesis 1a: The effect of asset specificity actions appropriate. Actions that have been used by
on the probability of choosing an equity (as best-practice firms in the host country become
opposed to non-equity) foreign market entry legitimized and are taken for granted in the eyes
mode is weakened in the presence of interna- of the firm’s stakeholders or of other foreign market
tional experience. actors (Cheng & Yu, 2008; Guillén, 2003; Scott,
1995). Legitimized actions will be less likely to
Host-country networks be questioned or challenged, making it irrational
We define host-country networks as the extent of for foreign market players or the firm’s stakeholders
an SME’s relevant business relationships with custo- to behave opportunistically. With lower threats
mers, business partners, and competitors in the of opportunistic behavior, internal uncertainty
target market (Coviello & Munro, 1997). We argue decreases, and hence SMEs will find it less necessary
that when host-country networks are present, to control their specific assets via internalization.
the need to internalize specific assets through the Hence SMEs can benefit from the knowledge of
choice of an equity mode lessens (Delios & Henisz, best-practice firms without necessarily having to
2000). The underlying rationale is consistent with build up the knowledge themselves.
arguments by scholars who have suggested that According to Shaver, Mitchell, and Yeung (1997),
relational exchange arrangements can complement firms can benefit from imitation only when they
or even substitute for formal contracts (Mesquita & exceed a critical threshold of individual experience
Brush, 2008; Poppo & Zenger, 2002). As host- in the host country.
country networks grow, two effects cause opportu-
Prior operations in the host country provide a general
nistic behavior by business partners to decline
understanding of the host country environment that
(Provan, 1993). First, networks influence the quality managers require in order to interpret information spil-
and flow of information and, hence, knowledge lovers. By contrast, firms without operations in the host
about the host country (Granovetter, 2005). country often are unaware of other firms’ actions in the
Research has shown that learning from networks is host country. (Shaver et al., 1997: 813)
particularly relevant for SMEs (Coviello & Munro,
These arguments may challenge our line of
1997; Schwens & Kabst, 2009). By providing infor-
reasoning, because SMEs often lack internationali-
mation, networks safeguard against internal uncer-
zation experience. However, imitation as defined
tainty, decreasing the need for high control modes
in the present paper focuses on a subset of organi-
despite high asset specificity. Second, when both
zations based on easily identifiable characteristics
business partners are members of the same net-
(i.e., success), facilitating SMEs to identify the
work, an internationalizing SME can benefit from
relevant actions to imitate (Cheng & Yu, 2008).
collective sanctions to protect against opportunistic
Furthermore, SMEs do not necessarily require
behavior by a partner (Granovetter, 2005). Hence
information spillovers to benefit from imitation.
host-country networks represent a viable mecha-
Instead, SMEs benefit from the legitimacy of best-
nism for reducing internal uncertainty, and can act
practice firms’ actions in the host country, without
as an alternative safeguard against opportunistic
necessarily acquiring the underlying rationale.
behavior for SMEs (Barkema, Bell, & Pennings, 1996;
Delios & Beamish, 1999; Ellis, 2000), making the
Hypothesis 1c: The effect of asset specificity on
costly internalization of specific assets less relevant.
the probability of choosing an equity (as opposed
to non-equity) foreign market entry mode is
Hypothesis 1b: The effect of asset specificity
weakened in the presence of imitation.
on the probability of choosing an equity (as
opposed to non-equity) foreign market entry
mode is weakened in the presence of host-
Institutional Safeguards
country networks.
Property rights protection
Imitation Property rights protection is the existence and
We define imitation as a mimetic behavior that enforcement of rules and regulations barring
places particular weight on companies in the host the expropriation of firm property (Henisz &

Journal of International Business Studies

 Asset specificity and SME entry mode Birger Maekelburger et al
463

Williamson, 1999). Low levels of property rights market (King & Shaver, 2001). Foreign firms are
protection create difficulties during the three stages subject to discrimination by host-country con-
of contracting: the specification of property rights, sumers, governments, and suppliers (Hymer,
monitoring, and enforcement (Oxley, 1999). The 1976). Furthermore, cultural differences result in
insufficient protection of property rights invites a higher complexity for strategic and organiza-
opportunistic behavior, and results in external tional decision-making (Shane, Venkataraman,
uncertainty (e.g., Agarwal & Ramaswami, 1992; & MacMillan, 1995). These disadvantages foster
Ahmed, Mohamad, Tan, & Johnson, 2002). The risk adverse behavior and increase the chance of
of opportunistic behavior is particularly severe failure, which makes contracting both difficult
when firms have specific assets to protect. One and costly.
means of overcoming opportunism is internali- Practice and theory suggest that some cultures
zation, which helps to reduce contractual hazards are closer than others, and internationalizing
by replacing some bilateral negotiations through firms “are more likely to fail whenever the accul-
internal control (Henisz, 2000b). However, suffi- turation involved is more demanding” (Barkema
cient institutional property rights protection faci- et al., 1996: 154). The lever determining the
litates contracting by decreasing the potential difficulty of acculturation is the distance between
for opportunistic behavior (Oxley, 1999). The need home- and host-country cultures. Choosing cul-
to protect specific assets through internalization turally proximal markets facilitates a firm’s accul-
thus declines. turation, and decreases the likelihood of adverse
Patents, copyrights, and trade secrets are means behavior and the risk of failure. Therefore contract-
of protecting specific assets, and can prevent both ing between similar cultures is easier and inter-
replication and the development of substitutes nalization less relevant, even when asset specificity
(Oviatt & McDougall, 1994). Although the pro- is high.
tection of assets is relevant for any type of firm, We believe that because SMEs are particularly
the argument is particularly important for SMEs, sensitive to external influences, smaller firms
because they are known to target niche markets are strongly affected by cultural distance. Child,
with specialized products and services (Nakos & Rodrigues, and Frynas (2009) argue that whereas
Brouthers, 2002). However, SMEs may not have larger firms may have sufficient bargaining power
the resources to apply for patents (Gassmann & to overcome cultural issues in a formal way, SMEs
Keupp, 2007), and patents may be ignored in some need to use more informal means or find ways
countries (Oviatt & McDougall, 1994). Conseque- to avoid the problem altogether to reduce the risk
ntly, one of the most important regulatory factors of adverse behavior. Because such results are often
in reducing external uncertainty and protec- difficult to achieve, venturing into culturally
ting specific SME assets is the existence of and proximal markets can act as a safeguard for SMEs.
adherence to property rights in the foreign market.
The risk of opportunistic behavior is thereby Hypothesis 2b: The effect of asset specificity on
reduced, and internalization is less critical even the probability of choosing an equity (as opposed
if asset specificity is high (Delios & Beamish, 1999). to non-equity) foreign market entry mode is
weakened in the presence of cultural proximity.
Hypothesis 2a: The effect of asset specificity
on the probability of choosing an equity (as
opposed to non-equity) foreign market entry METHODS
mode is weakened in the presence of property
rights protection. Data
Our empirical analysis draws on data gathered from
German SMEs. On the basis of the classifica-
Cultural proximity tion developed by the German Institute for SMEs
Cultural proximity is the extent to which socio- (Institut für Mittelstandsforschung, 2004), we
cultural factors such as values and operating define SMEs as companies with up to 500 employ-
methods are similar in the home and host countries ees. Furthermore, we selected only companies
(Gatignon & Anderson, 1988). More often than that had experienced some form of international
not, internationalizing firms are at a disadvan- activity. Using the established Hoppenstedt enter-
tage compared with domestic firms in the host prise database, we found 2549 companies that

Journal of International Business Studies

 Asset specificity and SME entry mode Birger Maekelburger et al
464

fulfilled both criteria. We sent standardized ques- items to measure the variables for our empirical
tionnaires to firm owners and CEOs, a target group model. To measure the central independent vari-
chosen because of its paramount importance in able asset specificity, we used a three-item scale
strategic decision-making in SMEs. This process (Cronbach’s a¼0.693) (Geyskens et al., 2006). The
was consistent with that used by Mesquita and introduction, heading the three relevant item
Lazzarini (2008: 365), who argued that “the CEO questions, asked respondents to describe what
of an SME is invariably the person who has the specific know-how and financial resources were
authority for all major decisions taken.” The sample transferred into the foreign venture that would be
included only German firms, and all questions of less value in other markets. The three items were
were in German. We took into consideration the “The foreign venture utilizes new firm-internal
established back-translation literature, and used technologies,” “Significant firm-internal know-how
internationally accepted items (Brislin, 1970; Hui was transferred into the foreign venture,” and
& Triandis, 1985; Van de Veijver & Hambleton, “The foreign venture required significant financial
1996). We received 257 completed questionnaires investments” (1¼low extent, 5¼high extent).
for a 10.1% response rate. Because of missing Our international experience measure includes
data, our final sample includes 206 companies. the level of international experience of the manage-
The average company size was 241 employees. ment (Bloodgood, Sapienza, & Almeida, 1996) and
We conducted tests as suggested by Little and that of the company as a whole (Burgel & Murray,
Rubin (1987) and Allison (2002) to account for 2000). We measured international experience using
missing data, but did not achieve any significant a three-item scale (Cronbach’s a¼0.818), and inclu-
results. Finally, we controlled for nonresponse ded statements such as “Before entering the foreign
bias following Armstrong and Overton (1977), market, the management had acquired long-lasting
comparing early and late respondents in terms of and far-reaching international experience” and
selected constructs. A t-test revealed no significant “Before starting this foreign venture, the company
differences. possessed considerable and long-lasting interna-
tional experience” (1¼low extent, 5¼high extent).
Measurements To measure host-country networks, we used a three-
We asked respondents to answer all questions item scale (Cronbach’s a¼0.703) (Ellis & Pecotich,
related to the SME’s entry into its principal foreign 2001). Sample items included “Knowledge about
market (Bradley & Gannon, 2000). For our depen- and contact with business partners in the foreign
dent variable foreign market entry mode, we followed market was well established” and “A comprehen-
Hennart (1988), Erramilli and D’Souza (1993), and sive network of business relationships with custo-
Nakos and Brouthers (2002). We distinguish bet- mers was already established in the foreign market”
ween equity foreign market entry modes (coded (1¼low extent, 5¼high extent).
as 1), including equity joint ventures and wholly We measured imitation using a three-item scale
owned subsidiaries, and non-equity foreign market (Cronbach’s a¼0.690) (Schwens & Kabst, 2009).
entry modes (coded as 0), including exports and Statements included were “Other companies in the
contractual agreements. In our sample of 206 foreign market were used as examples in the entry
German SMEs, we identified 88 companies (43%) mode selection process” and “Companies perceived
that had chosen equity foreign market entry modes as ‘best practice’ were used as examples in the
and 118 SMEs (57%) that had opted for non-equity entry mode selection process” (1¼low extent,
foreign market entry modes. Our sample contains 5¼high extent).
206 market entries into 26 different countries: 87 To measure property rights protection, we used
(42%) in Western Europe (e.g., Austria, Denmark, the established International Property Rights Index
France); 53 (26%) in North America (i.e., United (IPRI) (Dedigama, 2009). Of the three areas mea-
States, Canada, and Mexico); 35 (17%) in Eastern sured in the IPRI (Legal and Political Environment,
Europe (e.g., Hungary, Czech Republic, Poland); Physical Property Rights, and Intellectual Property
22 (11%) in Asia (e.g., Japan, China, Malaysia); and Rights (IPR)), we focused on IPR, because it corres-
the remaining 9 (4%) in Africa (Morocco, South ponds best with Williamson’s definition of property
Africa) and South America (Ecuador). The majority rights. IPR includes the variables protection of inte-
of firms hence opted for developed countries. llectual property rights, patent protection, and copyright
The independent variables were largely collected piracy, derived from World Bank and World
using a 5-point Likert scale. We adapted established Economic Forum data. All variables are measured

Journal of International Business Studies

 Asset specificity and SME entry mode Birger Maekelburger et al
465

on a scale between 0 (non-existence of property as control variables. The structure and impact of
rights) and 10 (strongest level of property rights). motives play a major role in internationalization
To measure the cultural proximity between home (Sarkar & Cavusgil, 1996; Williams, 1992). There-
and host countries, we used Kogut and Singh’s fore we included foreign market access, know-how
(1988) well-known construct of cultural distance. access, and cost reduction as internationalization
However, whereas the original index included motives.
Hofstede’s four cultural dimensions, we drew upon Because most of the measures used in our study
GLOBE’s nine cultural dimensions (House, Hanges, are self-reported and collected from an identical
Javidan, Dorfman, & Gupta, 2004; Javidan & House, source, common method variance (CMV) may be
2001). We used the GLOBE practices because the an issue (Hair, Anderson, Tatham, & Black, 1995).
entry mode choice is likely to be determined by However, we do not believe this to be a problem in
the situation as it is in the host country, rather than our study, for at least four reasons. First, our
by what it should be (as expressed by GLOBE dependent variable (equity vs non-equity foreign
values). GLOBE was used in this research because market entry mode) is objective in nature, rather
that project has moved beyond alternative cultural than based on perception, and hence does not
measures, designing constructs and scales that are create bias (Brouthers & Brouthers, 2003). Second,
more comprehensive, cross-culturally developed, we obtained two of our moderators (property rights
and theoretically sound. They have also been empi- protection and cultural proximity) from secondary
rically verified (Javidan, Dorfman, Luque, & sources, rather than relying on interviewee per-
House, 2006). We adapted the original formula by ceptions. Third, the inclusion of several interaction
Kogut and Singh (1988) and calculated our cultural terms is likely to reduce CMV, “because such
distance index accordingly: a complex relation is, in all likelihood, not part
of the respondents’ theory-in-use” (Chang, van
9 h
X 2 i
Witteloostuijn, & Eden, 2010: 180). Fourth, we
CDi ¼ Iij  Iig =Vi =9 ð1Þ
i¼1
used the one-factor test, as suggested by Podsakoff
and Organ (1986), and discovered no significant
where Iij represents the ith of nine cultural dimen- problems. When we loaded our variables, four
sions in the jth country; g represents Germany; Vi is factors were created, of which the largest explained
the variance of the index of the ith of nine only 22% of the variance.
dimensions; and CDi is the cultural distance
between Germany and the jth country (Kogut & Analysis and Results
Singh, 1988). To measure cultural proximity rather Table 1 presents the results of our multicollinearity
than cultural distance, we recoded the variable. analysis. Except for the high correlation between
We also included a set of control variables that property rights protection and cultural proximity,
may affect foreign market entry mode choice. all values stay well below 0.7 (Anderson, Sweeney,
Because host-country policies can influence foreign & Williams, 1996). This finding was confirmed
modes of entry, we included the POLCON index by calculating the variance inflation factors (VIFs)
to measure political constraints (Henisz, 2000a). to test the potential impact of collinearity. The
The advantage of this index is that the data not highest values are 2.41 for property rights pro-
only are available on a per-country basis, but also tection and 2.15 for cultural proximity, remaining
are available for many points in time, covering below the critical threshold of 2.5 (Allison, 1999).
changes in the political environment.2 The avai- Hence we do not assume multicollinearity to be
lability of resources plays a significant role in SMEs. a major problem.3
We therefore included firm size as a proxy for To analyze the validity of our constructs, we con-
resource availability. In line with prior studies (e.g., ducted principal component analysis with varimax
Rasheed, 2005), we used the log of the number rotation. We used 12 items to generate the four
of employees to measure firm size. Studying SMEs, components asset specificity, international experi-
we also decided to include family business as a ence, host-country networks, and imitation. Table 2
dichotomous control variable (1¼family business, shows the rotated component matrix obtained via
0¼non-family business). For SMEs, internationali- the principal component analysis, which indicated
zation is an important strategic decision that is in clean factor loadings for each item and construct.
turn often strongly influenced by family members Because our dependent variable, foreign market
(Basly, 2007). Finally, we included a set of motives entry mode, is dichotomous, we chose a binary

Journal of International Business Studies

 Asset specificity and SME entry mode Birger Maekelburger et al
466

Table 2 Results from principal component analysis

1.88
1.04
1.27
13

Items Component

0.17*
1.91
0.92
1.27
12

Asset International Host-country Imitation

specificity experience networks

0.21**
0.17*
3.57
0.72
1.39

0.021 0.018
11

Item 1 0.785 0.217

Item 2 0.827 0.075 0.234 0.095
Item 3 0.684 0.195 0.044 0.205
1.23
0.42
1.12

0.03
0.06
0.11
Item 4 0.081 0.771 0.198 0.121
10

Item 5 0.103 0.817 0.159 0.234

Item 6 0.075 0.858 0.233 0.088

0.19**
0.18*
Item 7 0.223 0.163 0.835 0.071
5.35
0.52
1.17

0.02
0.06
9

Item 8 0.230 0.214 0.798 0.057

Item 9 0.203 0.245 0.607 0.185
Item 10 0.034 0.171 0.039 0.781
0.37
0.16
1.38

0.00
0.10
0.11
0.01
0.02 Item 11 0.206 0.050 0.118 0.858
8

Item 12 0.040 0.191 0.256 0.597

0.37***

Note: Values highlighted in bold italic demonstrate the relationship

0.16*

0.14*
2.33
1.44
2.15

0.03
0.04

0.03

between items and factors.

7

0.72***
0.42***

logistic regression model for our analyses. We

0.20**
7.48
1.35
2.41

0.01
0.06

0.03
0.05

report the results of the regression analysis in

6

Table 3. To display the results of our logistic

regression analysis, we defined seven models. As
0.14*
0.14*
0.15*
2.31
0.72
1.23

0.06
0.04
0.06
0.08
0.05

proposed by Aiken and West (1991), establishing

5
Mean values, standard deviations, variance inflation factor values, and correlations

different models makes it possible to compare

0.30***

0.32***

alternative models by isolating changes in model

0.16*
2.32
0.77
1.62

0.06
0.09
0.06
0.09
0.08

0.13
4

fit and determining the explanatory power of the

variables. Model 1 includes the control variables
0.49***
0.36***

0.31***

and their effect on the foreign market entry mode.

0.18**
2.41
0.87
1.55

0.06
0.02
0.02

0.07
0.10
0.12

It shows that only the motive of cost reduction is

3

significantly and positively associated with equity

foreign market entry modes (p¼0.054). Firms seem
0.25***

0.33***
0.23**

to prefer equity modes to benefit from labor cost

0.15*

0.14*
0.14*

0.14*
2.10
0.85
1.37

0.04
0.02

0.03

0.14
2

advantages in the foreign market. Adding the

independent variable (asset specificity) and the
0.29***
0.31***
0.28***

moderator variables (international experience,

0.21**

0.16*
0.43
0.50

0.09
0.00
0.10
0.12
0.04
0.03
0.06

host-country networks, imitation, property rights

1

protection, and cultural proximity) in Model 2

significantly increases the explanatory power of
Company size (log no. of employees)

our model. Consistent with TCE-based arguing, we

find asset specificity to be a significant stimulus
Motive: Foreign market access

for equity entry modes (p¼0.033). Besides inter-

national experience (p¼0.027), most moderators
Motive: Know-how access
Property rights protection

*pp0.05; **pp0.01; ***pp0.001.

International experience

Motive: Cost reduction

do not have significant coefficients on the probabi-

Host-country networks
Standard deviations (s.d.)

lity of choosing equity entry modes. In Models 3–7,

Political constraints
Cultural proximity

we included the interaction effects. Before creating

Asset specificity

Family business

interaction terms, and to avoid multicollinearity,

Entry mode

we mean-centered all variables, as suggested by

Imitation
Mean values

Aiken and West (1991). Jaccard (2001) suggests

conducting a hierarchical test of changes in model
Variable
Table 1

fit to assess whether or not an interaction term

VIF

10
11
12
13
1
2
3
4
5
6
7
8
9

is trivial. Table 3 therefore reports the resulting

Journal of International Business Studies

 Asset specificity and SME entry mode Birger Maekelburger et al
467

Table 3 Logistic regression model: Foreign market entry mode (equity vs non-equity)

Variables Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Model 7

Independent variables
Asset specificity 0.43* 0.51* 0.53* 0.55* 0.57* 0.47*

Moderator variables
International experience 0.46* 0.39w 0.42w 0.49* 0.43* 0.46*
Host-country networks 0.34 0.33 0.40w 0.29 0.39w 0.40w
Imitation 0.29 0.31w 0.25 0.33w 0.40* 0.40*
Property rights protection 0.28 0.38 0.35 0.25 0.35 0.30
Cultural proximity 0.05 0.09 0.09 0.08 0.03 0.06

Interaction terms
Asset specificity  International experience 0.44*
Asset specificity  Host-country networks 0.55*
Asset specificity  Imitation 0.37w
Asset specificity  Property rights protection 0.50w
Asset specificity  Cultural proximity 0.45*

Control variables
Political constraints 1.35 1.15 1.12 1.45 0.79 1.11 0.90
Company size (log no. of employees) 0.45 0.17 0.28 0.27 0.16 0.09 0.15
Family business 0.02 0.14 0.14 0.21 0.14 0.16 0.25
Motive: Foreign market access 0.02 0.33 0.28 0.29 0.35 0.48w 0.44
Motive: Know-how access 0.09 0.14 0.16 0.17 0.20 0.18 0.20
Motive: Cost reduction 0.28w 0.03 0.06 0.06 0.01 0.04 0.02

Reliability
Model w2 9.52 39.51*** 43.91*** 46.24*** 43.40*** 43.51*** 43.98***
DModel w2 (vs Model 2) – – 4.40* 6.73** 3.89* 4.00* 4.47*
R2 (Cox & Snell) 0.05 0.19 0.21 0.22 0.20 0.20 0.21
R2 (Nagelkerke) 0.06 0.25 0.28 0.29 0.27 0.28 0.28
Correct classification (%) 61.5 67.5 69.1 71.2 68.6 68.6 68.1
*pp0.05; **pp0.01; ***pp0.001; wpp0.10.

Dw2 values for Models 3–7 in comparison with action terms. The commands report differences in
that of Model 2. All Dw2 values are significant, the predicted probabilities associated with discrete
which suggests that none of the interaction terms is changes in key independent variable values. The
trivial. advantage of this method is the use of a simulation-
Although previous studies have often compared based approach to statistically interpret the inter-
changes in model fit after adding interaction terms, action based on confidence intervals for different
this process does not mitigate the problems asso- values of the independent variable. The method,
ciated with detecting and interpreting interaction as suggested by Zelner (2009), results in the two
effects (Hoetker, 2007). Interactions in nonlinear sets of plots in Figures 2–6. The first set of plots
models compared with linear models are more shows the predicted probabilities of equity foreign
difficult to analyze, as they cannot be interpreted market entry modes (y-axis) at different levels of
by simply examining the direction, magnitude, asset specificity (x-axis) for high and low values of
and statistical significance of the resulting coeffi- the moderator variables (dashed and solid line).
cients. The direction of the interaction term and Although these plots are commonly used, Zelner
the statistical significance may vary for different (2009: 1338) argues that
values of the independent variable (Hoetker, 2007;
they do not by themselves provide a sufficient basis on
Norton, Wang, & Ai, 2004; Zelner, 2009). Zelner which to draw sound statistical conclusions. The reason is,
(2009) suggests using a set of Stata commands to like the estimated coefficients on which they are based,
assess the direction and significance of the inter- predicted probabilities are also estimates.

Journal of International Business Studies

 Asset specificity and SME entry mode Birger Maekelburger et al
468

1 1

(Equity Foreign Market Entry Mode)

(Equity Foreign Market Entry Mode)

0.8 0.8

Predicted Probability
Predicted Probability

0.6 0.6 B

B
0.4 0.4

0.2 0.2

0 0
low medium high low medium high
Asset Specificity Asset Specificity
90% Confidence Interval Low International Experience 90% Confidence Interval Low Host-Country Networks
90% Confidence Interval High International Experience 90% Confidence Interval High Host-Country Networks

0.4
(Equity Foreign Market Entry Mode)

0.2 0.4

(Equity Foreign Market Entry Mode)

B C 0.2
Δ Predicted Probabilities

Δ Predicted Probabilities
B C
A 0
A
-0.2
-0.2
-0.4
-0.4
-0.6 -0.6

-0.8 -0.8

-1 -1

low medium high low medium high

Asset Specificity Asset Specificity

90% Confidence Interval High vs. Low 90% Confidence Interval High vs. Low
International Experience Host-Country Networks

Figure 2 Interaction plots: international experience. Figure 3 Interaction plots: host-country networks.
(a) Predicted probabilities for high and low international (a) Predicted probabilities for high and low host-country
experience; (b) Delta predicted probabilities for high vs low networks; (b) Delta predicted probabilities for high vs low
international experience. host-country networks.

Therefore we include a second set of plots that support Hypothesis 1a, in which we suggest that
represent the difference in predicted probabilities international experience weakens the effect of asset
between high and low values of the moderator specificity on the choice of equity entry modes. For
variables for different values of asset specificity (i.e., medium values of asset specificity, the interaction is
the vertical distance between the dashed and solid not significant (the confidence intervals include
lines in the first set of plots), including the 90% zero). Interestingly, the moderator is positive and
confidence intervals for these differences for differ- significant for low values of asset specificity (to the
ent values of asset specificity. left of Point A), which indicates that firms with
The significant interaction terms of the knowl- high levels of international experience are more
edge safeguards included in Models 3–5 indicate likely to choose equity entry modes when asset
that international experience (p¼0.041), host-coun- specificity is low, hinting at the direct effect of
try networks (p¼0.013), and imitation (p¼0.056) international experience on equity entry mode.
are significant and negative moderators. Figure 2a Figures 3a and 3b provide a graphical analysis of
demonstrates that at high levels of international the host-country networks moderator. The plots are
experience (dashed line) the effect of asset specifi- similar to the international experience plots. The
city on equity foreign market entry mode is weak significant positive effect at low levels of asset
(illustrated by the almost horizontal dashed line). specificity (to the left of Point A) indicates that a
Figure 2b shows that international experience is high degree of host-country networks increases
a negative moderator from low to medium values an SME’s tendency to choose equity entry modes
of asset specificity (to the right of Point B). At when asset specificity is low. Whereas international
a very high level of asset specificity (Point C), the experience is significant only at a very high level
moderator is statistically significant (the confi- of asset specificity, host-country networks is a
dence interval does not include zero). These results significant negative moderator from medium to

Journal of International Business Studies

 Asset specificity and SME entry mode Birger Maekelburger et al
469

1 1

(Equity Foreign Market Entry Mode)

(Equity Foreign Market Entry Mode)

0.8 0.8

Predicted Probability
Predicted Probability

0.6 0.6

B
0.4 0.4
B

0.2 0.2

0 0
low medium high low medium high
Asset Specificity Asset Specificity

90% Confidence Interval Low Imitation 90% Confidence Interval Low Property Rights Protection
90% Confidence Interval High Imitation 90% Confidence Interval High Property Rights Protection

0.4 0.4

(Equity Foreign Market Entry Mode)

(Equity Foreign Market Entry Mode)

0.2 0.2
B B

Δ Predicted Probabilities
C C
Δ Predicted Probabilities

0 0
A
-0.2 -0.2

-0.4 -0.4

-0.6 -0.6

-0.8 -0.8

-1 -1
low medium high low medium high
Asset Specificity Asset Specificity

90% Confidence Interval High vs. Low Imitation 90% Confidence Interval High vs. Low Property Rights Protection

Figure 4 Interaction plots: imitation. Figure 5 Interaction plots: property rights protection.
(a) Predicted probabilities for high and low imitation; (a) Predicted probabilities for high and low property rights
(b) Delta predicted probabilities for high vs low imitation. protection; (b) Delta predicted probabilities for high vs low
property rights protection.

high values (to the right of Point C). These results specificity and equity foreign market entry modes.
support Hypothesis 1b, in which we suggest that Figures 5a and 5b suggest that, from a low level of
host-country networks weaken the relationship asset specificity (to the right of Point B), property
between asset specificity and an SME’s propensity rights protection negatively moderates the relation-
to choose equity entry modes. ship between asset specificity and equity entry
Figures 4a and 4b provide graphical evidence modes. The interaction is significant from medium
for the influence of the imitation moderator. The to high values of asset specificity (to the right of
plots provide results similar to those for the other Point C), supporting Hypothesis 2a. Even though
two knowledge safeguards. Imitation becomes a the moderating effect of property rights protection
negative moderator at low to medium values of is positive at a very low level of asset specificity (to
asset specificity (to the right of Point B), and is the left of Point B), this effect is not statistically
significant from medium to high values of asset significant (the confidence interval includes zero),
specificity (to the right of Point C), supporting which suggests that property rights protection has a
Hypothesis 1c. As was the case with the two pre- limited effect on an SME’s propensity to choose
vious plots, imitation is a significant positive equity entry modes at low levels of asset specificity.
moderator at low values of asset specificity (to the Finally, Figures 6a and 6b suggest that cultural
left of Point A), which demonstrates that imitation proximity becomes a significant and negative
increases an SME’s propensity to choose equity moderator of the relationship between asset speci-
entry modes when asset specificity is low. ficity and entry modes at a medium to high level of
The interaction terms for the institutional safe- asset specificity (to the right of Point C). These
guards as presented in Models 6 and 7 are also findings support Hypothesis 2b. Furthermore, at a
statistically significant, suggesting that property very low level of asset specificity (Point A), cultural
rights protection (p¼0.062) and cultural proximity proximity becomes a significant and positive
(p¼0.045) moderate the relationship between asset moderator, suggesting that cultural proximity has

Journal of International Business Studies

 Asset specificity and SME entry mode Birger Maekelburger et al
470

1
between institutional factors and forms of asset
(Equity Foreign Market Entry Mode)

0.8 specificity (e.g., Dikova & van Witteloostuijn, 2007;

Erramilli & Rao, 1993; Schwens et al., 2011), which
Predicted Probability

0.6
links institutional theory and TCE-based argu-
0.4
B
ments. These studies have attempted to support
TCE’s original suggestion that asset specificity is
0.2 problematic when coupled with uncertainty. Very
0
few studies, however, have linked asset specificity,
low medium high knowledge factors, and institutional factors as in
Asset Specificity the present study.
90% Confidence Interval Low Cultural Proximity The link between TCE and knowledge safeguards
90% Confidence Interval High Cultural Proximity
is internal uncertainty. Delios and Henisz (2000)
0.4
tested the moderating influence of both industry
and national experience on the relationship
(Equity Foreign Market Entry Mode)

0.2
between private expropriation hazards (defined as
Δ Predicted Probabilities

C
0
A B R&D intensity, a measure often used for asset
-0.2
specificity) and the propensity to choose equity
-0.4 ownership. The findings confirmed that inexper-
-0.6 ienced firms are much more sensitive to changes in
-0.8 private expropriation hazards than are experienced
-1 firms. The study results are consistent with our
low medium high findings regarding the interaction between inter-
Asset Specificity
national experience and asset specificity. We
90% Confidence Interval High vs. Low Cultural Proximity
further expand this knowledge by introducing the
Figure 6 Interaction plots: cultural proximity. interplay between asset specificity and the other
(a) Predicted probabilities for high and low cultural proximity; two knowledge safeguards – host-country networks
(b) Delta predicted probabilities for high vs low cultural and imitation – thereby providing more com-
proximity. prehensive findings. We encourage researchers to
employ models that take into account the inter-
dependencies among key factors to further con-
textualize foreign market mode choice and its
a direct positive effect on equity foreign market antecedents.
entry mode choice. The link between TCE and institutional safe-
guards is external uncertainty. Dikova and van
DISCUSSION Witteloostuijn (2007) could not find support for
In this section, we discuss our findings in light of their hypothesis that institutional advancement
previous research. Furthermore, we speculate on moderates the relationship between technological
how the SME-specific ideas developed in the intensity and the tendency to share ownership.
present paper can inform TCE-based research on Their sample primarily contained larger firms (56%
foreign market entry mode choice and beyond. over 1200 employees), which we believe to be the
reason for the difference between their results and
Discussion in Light of Previous Research our findings. Studying the foreign market entry
Forging links between complementary research mode decisions of service firms, Erramilli and Rao
streams has significantly increased our understand- (1993) could show a low significant relationship
ing of the foreign market entry mode decisions of between country risk and asset specificity. However,
both SMEs and large MNEs. One of the most com- the significance seems to have arisen primarily
monly investigated interactions in the entry mode from the preference of low-specificity firms for
literature is that between knowledge and insti- non-equity entry modes in high-risk situations.
tutional factors (e.g., Cho & Padmanabhan, 2005; Although our results similarly suggest that low-
Delios & Henisz, 2000; Henisz & Delios, 2001). specificity firms prefer non-equity entry modes, we
Studies have shown that knowledge can reduce cannot find evidence for differences caused by
perceived environmental uncertainty in a host property rights protection when asset specificity is
country. Another commonly studied interaction is low. In fact, our results suggest that the interaction

Journal of International Business Studies

 Asset specificity and SME entry mode Birger Maekelburger et al
471

is caused largely by high-specificity firms that prefer experienced by smaller firms, they can affect
equity entry modes in situations involving low large MNEs under certain conditions as well.
property rights protection. Because our results Furthermore, large MNEs face typical liabilities
are in line with Li and Qian’s (2008) conclusions such as organizational inertia (Hannan & Freeman,
indicating that SMEs prefer internalization when 1984), inefficient knowledge flows (Kostova, 1999),
their specific assets are exposed to a high risk of and obstacles associated with geographic dispersion
asset leakage, we believe that firm size is the key (Mentzer, 1993). Thus, the alternative safeguarding
explanatory factor influencing these differences; rationale discussed in the present paper can inform
approximately 70% of the service firms observed by TCE-based international mode choice research
Erramilli and Rao (1993) were large MNEs. Further- on large MNEs as well. One specific example might
more, Erramilli and Rao (1993) did not find be post-entry-mode decisions of large MNEs. For
significant evidence for the interaction between instance, when it would be advisable for large MNEs
asset specificity and cultural distance. They suggest to change their entry modes in a foreign market
that the reason for the insignificant findings is that due to changes in institutional settings, liabilities
cultural distance does not raise transaction costs such as organizational inertia may hamper the
sufficiently to impact on the relationship between adaptation process. Drawing on TCE and institu-
asset specificity and entry mode choice. In contrast, tional theory, Puck, Holtbrügge, and Mohr (2009)
our paper reveals cultural proximity to be a signi- investigate mode change by large MNEs in China
ficant moderator of the relationship between asset and hypothesize a direct positive effect of asset
specificity and entry modes. Once again, we believe specificity on the likelihood of a conversion of joint
that the divergent findings are due to our focus ventures into wholly owned subsidiaries. However,
on SMEs, which are more sensitive to external they do not find support for this relationship.
influences, as our results are consistent with According to Puck et al. (2009: 397, no italics
Erramilli and D’Souza’s (1993) findings indicating in original), one possible explanation is that
that small firms prefer similar markets when assets MNEs “may have developed safeguards against the
are at stake. Schwens et al. (2011) investigated unintended dissipation” of their specific assets.
SMEs and found both formal institutional risk and These alternative safeguarding mechanisms might
informal institutional distance to exert a moderat- include the experience gathered and network
ing influence on the relationship between proprie- contacts developed in the host country that safe-
tary know-how and entry mode choice. Although guard the firm from opportunistic behavior, mak-
the authors investigated formal risk and informal ing internalization and expensive organizational
distance factors and we investigated formal and changes less relevant. Hence we encourage future
informal safeguards, the results from our study are TCE-based foreign market mode choice research
congruent with the former findings. This consis- to test theoretical predictions related to alter-
tency supports our suggestion that institutional native safeguarding mechanisms in the context of
safeguards are of particular pertinence for SMEs. large MNEs.
Furthermore, we believe that some of the insights
Discussion Beyond SMEs gleaned from the present research may have the
TCE is an important theoretical foundation in potential to inform TCE-based research beyond
foreign market entry mode research and beyond. foreign market mode choices. In a recent meta-
Brouthers and Hennart (2007: 400) note that TCE analysis, Geyskens et al. (2006) quantitatively inte-
“is the most widely used theoretical perspective in grated and evaluated results from 200 TCE-based
international entry mode research and appears empirical studies. From their comprehensive study
in almost half of the studies.” Moreover, “few results, Geyskens et al. (2006: 532) contend that
organizational frameworks have been studied for
a longer period of time or have been accorded empirical investigations of asset specificity in the future will
as much scholarly attention as transaction cost fruitfully deepen researchers’ understanding only if they go
theory” (Geyskens et al., 2006: 519). beyond simply demonstrating a main effect in the more
complex unresolved questions such as distinguishing
We specifically tailored the theoretical arguments
between transaction cost reasoning and alternative argu-
developed in the present paper to the liabilities ments.
of smallness: limited resources, lack of internatio-
nalization knowledge, and sensitivity to external The alternative safeguarding mechanisms identi-
influences. Although these liabilities are typically fied in the present paper may be a starting point

Journal of International Business Studies

 Asset specificity and SME entry mode Birger Maekelburger et al
472

for the further contextualization of TCE-based enterprise database to identify the relevant firms,
research, as it is necessary to “move transaction our acceptable response rate, and the results of our
cost theory beyond a dyadic focus” (Geyskens tests for nonresponse bias and outliers suggest that
et al., 2006: 533). Discussing their meta-analytical our data are representative.
findings, Geyskens et al. (2006) speculate on how Incorporating network theory into our research
social mechanisms and collective sanctions help without elaborating on the type of network, the
to guard against opportunistic behavior. This argu- position of the company within the network or the
ment is consistent with the arguments developed stability of the network relationships may seem
in the present paper. insufficient. For instance, we discuss networks
as both information providers and sanctioning
LIMITATIONS AND IMPLICATIONS mechanisms. However, although all networks of
As with most empirical studies, our study has which the focal company is a member can provide
limitations. One might argue that the level of asset information, only networks in which both business
specificity is not exogenously given, but is rather partners are members can have a sanctioning effect.
a matter of firm choice, and may reflect other Future research should supplement our research
underlying effects. Masten (1995: 60) suggests that model with more in-depth intra-network analysis
“the specificity of assets and the level of investment (Burt, 1992) or social-capital reasoning (Coleman,
in those assets that determine the size of appropri- 1988).
able quasi-rents are themselves decision variables.” The core indices chosen to reflect our institu-
Firms may choose to transfer considerable specific tional safeguards may be cause for some concern.
knowledge and funds into the foreign market only Indices measuring cultural distance assume that the
when it is reasonably certain that they can guard average of a country is an appropriate measure for
against opportunism. Although we cannot fully use in evaluating internationalization (Shenkar,
rule out endogeneity, we do not believe this to be 2001). However, countries may vary internally to
a major concern in our paper, for at least three such an extent that using Kogut and Singh’s (1988)
reasons. method might limit the explanatory power of the
First, our analysis might face the problem of research (Meyer & Nguyen, 2005). Because the
reverse causality. In such a case, not only would GLOBE dimensions (House et al., 2004; Javidan &
asset specificity affect foreign market entry modes, House, 2001) build on established indices such as
but entry modes would also have an effect on the the World Value Survey, we consider the resulting
level of asset specificity. Our implicit causality index to be a good proxy for management and
assumption that asset specificity determines the organizational issues. Our second index, the IPRI,
level of equity is consistent with prior TCE-based might come under similar critical scrutiny. Because
meta-analyses (Geyskens et al., 2006; Zhao et al., the IPRI is derived from established World Bank and
2004) and longitudinal studies that provided World Economic Forum data, and is collated from
evidence of the relationship between firm capabi- a distinct set of variables (Dedigama, 2009), we
lities and entry mode choice (Bogner, Thomas, & consider it a good representation for our purposes.
McGee, 1996). Although we do not believe reverse Our research focuses on companies from one
causality to be a major problem, longitudinal entry home country entering multiple host countries.
mode studies should be used to address this issue Germany has some of the strictest property rights
further in the future (Zhao et al., 2004). laws in the world, and for various cultural dimen-
Second, institutional factors are key sources or sions Germany is at the highest or lowest end.
indicators of opportunism. Hence asset specificity Further research should analyze knowledge and
may be determined by such factors. We tested institutional safeguards in other contexts.
the effect of the three institutional factors used in Our findings have implications for both SME
our analysis (property rights protection, cultural management and international chambers of com-
proximity, and political constraints) on asset speci- merce. We show that, despite resource constraints,
ficity by conducting a regression analysis but did SMEs tend to prefer equity modes, and accept the
not obtain significant results. associated risks when asset specificity is high.
Third, sample selection could be a cause of However, our study also shows that knowledge
endogeneity if our sample was non-representative can be a viable alternative safeguard, reducing
of SMEs with international activities in Germany. the propensity to select expensive equity modes.
However, our use of the established Hoppenstedt Managers with individual international experience

Journal of International Business Studies

 Asset specificity and SME entry mode Birger Maekelburger et al
473

or host-country networks can utilize these assets to the basis of the desired knowledge acquisition rather
protect specific investments. Although it may be than the amount of existing knowledge. We assume
difficult for managers to build up international this relationship to be less relevant for SMEs. First,
experience when needed, international chambers resource constraints limit SMEs in their ability to choose
of commerce may furnish network partners in equity entry modes that promise future knowledge
target countries. Imitation also appears to be a gains (Calof, 1994). Second, SMEs, as opposed to
viable method of overcoming internal uncertainty. large MNEs, often possess learning advantages per
Although blindly copying other companies may se, which facilitate knowledge acquisition irrespective
be dangerous, imitating successful companies of foreign market entry mode (Autio, Sapienza, &
can indeed be beneficial and legitimize the Almeida, 2000).
2
firm’s actions. Our study has shown that SMEs We also controlled for other country effects, using
feel significantly more comfortable in countries established indices such as the Economic Freedom
with strong property rights protection whose Index (Meyer, Estrin, Bhaumik, & Peng, 2009),
cultures are similar to those of their respective including the dimensions investment freedom, trade
home countries. Although these are factors that freedom, and business freedom. Because of high
the internationalizing company cannot influence correlations with the other country variables (i.e.,
directly, institutions should provide relevant advice political constraints, cultural proximity, and property
to properly support management decisions prior rights protection), we decided against including these
to choosing a foreign market. indices in the final model. However, despite the high
correlations, the key findings of our empirical analysis
ACKNOWLEDGEMENTS proved to be robust.
3
We thank Professor Gabriel Szulanski and three We also conducted a robustness check to deter-
anonymous reviewers for their constructive comments mine whether the correlation between cultural
and encouragement throughout the review process. proximity and property rights protection changes
our findings. We reran our models as illustrated in
NOTES Table 3 while excluding property rights protection
1
Mayer (2006) suggests that although knowledge and cultural proximity (both consecutively and
influences governance choices, governance choices simultaneously). The results were unchanged, and
also impact future knowledge gains. As a conse- hence the level of support for our hypotheses
quence, firms may make governance decisions on remained the same.

REFERENCES
Agarwal, S., & Ramaswami, S. N. 1992. Choice of foreign Basly, S. 2007. The internationalization of family SME: An orga-
market entry mode: Impact of ownership, location and nizational learning and knowledge development perspective.
internalization factors. Journal of International Business Studies, Baltic Journal of Management, 2(2): 154–180.
23(1): 1–27. Bloodgood, J. M., Sapienza, H. J., & Almeida, J. G. 1996. The inter-
Ahmed, Z. U., Mohamad, O., Tan, B., & Johnson, J. P. 2002. nationalization of new high-potential US ventures: Antecedents
International risk perceptions and mode of entry: A case study and outcomes. Entrepreneurship Theory & Practice, 20(4): 61–76.
of Malaysian multinational firms. Journal of Business Research, Bogner, W. C., Thomas, H., & McGee, J. 1996. A longitudinal
55(10): 805–813. study of the competitive positions and entry paths of European
Aiken, L., & West, S. 1991. Multiple regression: Testing and firms in the US pharmaceutical market. Strategic Management
interpreting interactions. Newbury Park, CA: Sage. Journal, 17(2): 85–107.
Allison, P. D. 1999. Multiple regression: A primer. Thousand Oaks, Bradley, F., & Gannon, M. 2000. Does the firm’s technology and
CA: Pine Forge Press. marketing profile affect foreign market entry? Journal of
Allison, P. D. 2002. Missing data. Thousand Oaks, CA: Sage. International Marketing, 8(4): 12–36.
Anderson, D. R., Sweeney, D. J., & Williams, T. A. 1996. Statistics Brislin, R. W. 1970. Back-translation for cross-cultural research.
for business and economics. Mason, OH: Thomson Higher Journal of Cross-Cultural Psychology, 1(3): 185–216.
Education. Brouthers, K. D., & Brouthers, L. E. 2003. Why service and
Anderson, E., & Gatignon, H. 1986. Modes of foreign entry: manufacturing entry mode choices differ: The influence of
A transaction cost analysis and propositions. Journal of transaction cost factors, risk and trust. Journal of Management
International Business Studies, 17(3): 1–26. Studies, 40(5): 1179–1204.
Armstrong, J. S., & Overton, T. S. 1977. Estimating nonres- Brouthers, K. D., & Hennart, J.-F. 2007. Boundaries of the firm:
ponse bias in mail surveys. Journal of Marketing Research, Insights from international entry mode research. Journal of
14(3): 396–402. Management, 33(3): 395–425.
Autio, E., Sapienza, H. J., & Almeida, J. G. 2000. Effects of age at Brouthers, K. D., & Nakos, G. 2004. SME entry mode choice and
entry, knowledge intensity, and imitability on international performance: A transaction cost perspective. Entrepreneurship
growth. Academy of Management Journal, 43(5): 909–924. Theory and Practice, 28(3): 229–247.
Barkema, H. G., Bell, J. H. J., & Pennings, J. M. 1996. Foreign Buckley, P. J. 1989. Foreign direct investment by small
entry, cultural barriers, and learning. Strategic Management and medium sized enterprises: The theoretical background.
Journal, 17(2): 151–166. Small Business Economics, 1(2): 89–100.

Journal of International Business Studies

 Asset specificity and SME entry mode Birger Maekelburger et al
474

Burgel, O., & Murray, G. C. 2000. The international market Granovetter, M. 2005. The impact of social structure on
entry choices of start-up companies in high-technology economic outcomes. Journal of Economic Perspectives, 19(1):
industries. Journal of International Marketing, 8(2): 33–62. 33–50.
Burt, R. S. 1992. Structural holes: The social structure of Guillén, M. F. 2003. Experience, imitation, and the sequence
competition. Cambridge, MA: Harvard University Press. of foreign entry: Wholly owned and joint-venture manufac-
Calof, J. L. 1994. The relationship between firm size and export turing by South Korean firms and business groups in China,
behavior revisited. Journal of International Business Studies, 1987–1995. Journal of International Business Studies, 34(2):
25(2): 367–387. 185–198.
Chang, S. J., van Witteloostuijn, A., & Eden, L. 2010. From Hair, J. F., Anderson, R. E., Tatham, R. L., & Black, W. C. 1995.
the editors: Common method variance in international Multivariate data analysis: With readings. Englewood Cliffs, NJ:
business research. Journal of International Business Studies, Prentice Hall.
41(2): 178–184. Hannan, M. T., & Freeman, J. 1984. Structural inertia and
Cheng, H.-L., & Yu, C.-M. J. 2008. Institutional pressures and organizational change. American Sociological Review, 49(2):
initiation of internationalization: Evidence from Taiwanese 149–164.
small- and medium-sized enterprises. International Business Henisz, W. J. 2000a. The institutional environment for economic
Review, 17(3): 331–348. growth. Economics and Politics, 12(1): 1–31.
Child, J., Rodrigues, S. B., & Frynas, J. G. 2009. Psychic distance, Henisz, W. J. 2000b. The institutional environment for multi-
its impact and coping modes. Management International national investment. Journal of Law, Economics, and Organiza-
Review, 49(2): 199–224. tion, 16(2): 334–464.
Cho, K. R., & Padmanabhan, P. 2005. Revisiting the role of Henisz, W. J., & Delios, A. 2001. Uncertainty, imitation,
cultural distance in MNC’s foreign ownership mode choice: and plant location: Japanese multinational corporations,
The moderating effect of experience attributes. International 1990–1996. Administrative Science Quarterly, 46(3): 443–475.
Business Review, 14(3): 307–324. Henisz, W. J., & Delios, A. 2002. Learning about the institutional
Coleman, J. S. 1988. Social capital in the creation of human environment. In P. Ingram & B. Silverman (Eds), The new
capital. American Journal of Sociology, 94(S1): 95–120. institutionalism in strategic management: 339–372. Advances in
Coviello, N., & Munro, H. 1997. Network relationships and International Management New York: JAI Press.
the internationalisation process of small software firms. Henisz, W. J., & Williamson, O. E. 1999. Comparative economic
International Business Review, 6(4): 361–386. organization: Within and between countries. Business and
Dedigama, A. C. 2009. International property rights index (IPRI) – Politics, 1(3): 261–278.
2009 report. Washington, DC: Property Rights Alliance. Hennart, J.-F. 1988. A transaction costs theory of equity joint
Delios, A., & Beamish, P. W. 1999. Ownership strategy of Japanese ventures. Strategic Management Journal, 9(4): 361–374.
firms: Transactional, institutional, and experience influences. Hennart, J.-F. 1991. The transaction costs theory of joint
Strategic Management Journal, 20(10): 915–933. ventures: An empirical study of Japanese subsidiaries in the
Delios, A., & Henisz, W. J. 2000. Japanese firms’ investment United States. Management Science, 37(4): 483–497.
strategies in emerging economies. Academy of Management Hesterly, W. S., & Zenger, T. R. 1993. Crossroads – the myth of a
Journal, 43(3): 305–323. monolithic economics: Fundamental assumptions and the use
Dikova, D., & van Witteloostuijn, A. 2007. Foreign direct of economic models in policy and strategy research. Organiza-
investment mode choice: Entry and establishment modes in tion Science, 4(3): 496–510.
transition economies. Journal of International Business Studies, Hoetker, G. 2007. The use of logit and probit models in strategic
38(6): 1013–1033. management research: Critical issues. Strategic Management
Ellis, P. 2000. Social ties and foreign market entry. Journal of Journal, 28(4): 331–343.
International Business Studies, 31(3): 443–469. House, R. J., Hanges, P. J., Javidan, M., Dorfman, P. W., & Gupta,
Ellis, P., & Pecotich, A. 2001. Social factors influencing export V. 2004. Culture, leadership, and organizations: The GLOBE
initiation in small and medium-sized enterprises. Journal of study of 62 societies. Thousand Oaks, CA: Sage.
Marketing Research, 38(1): 119–130. Hui, C. H., & Triandis, H. C. 1985. Measurement in cross-cultural
Erramilli, M. K., & D’Souza, D. E. 1993. Venturing into foreign psychology: A review and comparison of strategies. Journal of
markets: The case of the small service firm. Entrepreneurship Cross-Cultural Psychology, 16(2): 131–152.
Theory & Practice, 17(4): 29–41. Hymer, S. H. 1976. The international operations of national firms:
Erramilli, M. K., & D’Souza, D. E. 1995. Uncertainty and foreign A study of direct foreign investment. Cambridge, MA: The MIT
direct investment: The role of moderators. International Press.
Marketing Review, 12(3): 47–60. Institut für Mittelstandsforschung. 2004. SMEs in Germany: Facts
Erramilli, M. K., & Rao, C. P. 1993. Service firms’ international and figures 2004. Bonn: Institut für Mittelstandsforschung.
entry-mode choice: A modified transaction-cost analysis Jaccard, J. 2001. Interaction effects in logistic regression. Thousand
approach. The Journal of Marketing, 57(3): 19–38. Oaks, CA: Sage.
Forsgren, M. 2002. The concept of learning in the Uppsala Javidan, M., & House, R. J. 2001. Cultural acumen for the global
internationalization process model: A critical review. Interna- manager: Lessons from Project GLOBE. Organizational
tional Business Review, 11(3): 257–277. Dynamics, 29(4): 289–305.
Gassmann, O., & Keupp, M. M. 2007. The competitive Javidan, M., Dorfman, P. W., Luque, M. S., & House, R. J. 2006.
advantage of early and rapidly internationalising SMEs in the In the eye of the beholder: Cross cultural lessons in leadership
biotechnology industry: A knowledge-based view. Journal of from Project GLOBE. The Academy of Management Perspectives,
World Business, 42(3): 350–366. 20(1): 67–90.
Gatignon, H., & Anderson, E. 1988. The multinational corpor- King, A. A., & Shaver, J. M. 2001. Are aliens green?
ation’s degree of control over foreign subsidiaries: An Assessing foreign establishments’ environmental conduct
empirical test of a transaction cost explanation. Journal of in the United States. Strategic Management Journal, 22(11):
Law, Economics, and Organization, 4(2): 305–336. 1069–1085.
Geyskens, I., Steenkamp, J.-B. E. M., & Kumar, N. 2006. Make, Kogut, B., & Singh, H. 1988. The effect of national culture on
buy, or ally: A transaction cost theory meta-analysis. Academy the choice of entry mode. Journal of International Business
of Management Journal, 49(3): 519–543. Studies, 19(3): 411–432.
Granovetter, M. 1985. Economic action and social structure: The Kostova, T. 1999. Transnational transfer of strategic organiza-
problem of embeddedness. American Journal of Sociology, tional practices: A contextual perspective. Academy of
91(3): 481–510. Management Review, 24(2): 308–324.

Journal of International Business Studies

 Asset specificity and SME entry mode Birger Maekelburger et al
475

Li, L., & Qian, G. 2008. Partnership or self-reliance entry modes: internationalization on firm survival and growth. Academy of
Large and small technology-based enterprises’ strategies in Management Review, 31(4): 914–933.
overseas markets. Journal of International Entrepreneurship, Sarkar, M., & Cavusgil, S. T. 1996. Trends in international
6(4): 188–208. business thought and literature: A review of international
Little, R. J. A., & Rubin, D. B. 1987. Statistical analysis with market entry mode research: Integration and synthesis. The
missing data. New York: Wiley. International Executive, 38(6): 825–847.
Lu, J. W., & Beamish, P. W. 2001. The internationalization Schwens, C., & Kabst, R. 2009. How early opposed to late
and performance of SMEs. Strategic Management Journal, internationalizers learn: Experience of others and paradigms of
22(6/7): 565–586. interpretation. International Business Review, 18(5): 509–522.
Masten, S. E. 1995. Empirical research in transaction cost Schwens, C., Eiche, J., & Kabst, R. 2011. The moderating impact
economics: Challenges, progress, directions. In J. Groenewegen of informal institutional distance and formal institutional risk
(Ed), Transaction cost economics and beyond: 43–64. Amsterdam: on SME entry mode choice. Journal of Management Studies,
Kluwer. 48(2): 330–351.
Mayer, K. J. 2006. Spillovers and governance: An analysis Scott, W. R. 1995. Institutions and organizations. Thousand Oaks,
of knowledge and reputational spillovers in information CA: Sage.
technology. Academy of Management Journal, 49(1): 69–84. Shane, S., Venkataraman, S., & MacMillan, I. 1995. Cultural
Mayer, K. J., & Salomon, R. M. 2006. Capabilities, contractual differences in innovation championing strategies. Journal of
hazards, and governance: Integrating resource-based and Management, 21(5): 931–952.
transaction cost perspectives. Academy of Management Journal, Shaver, J. M., Mitchell, W., & Yeung, B. 1997. The effect of
49(5): 942–959. own-firm and other-firm experience on foreign direct invest-
Mentzer, M. S. 1993. The limits of control in the multinational ment survival in the United States, 1987–92. Strategic
corporation. International Journal of Commerce and Manage- Management Journal, 18(10): 811–824.
ment, 1(1/2): 55–72. Shenkar, O. 2001. Cultural distance revisited: Towards a more
Mesquita, L. F., & Brush, T. H. 2008. Untangling safeguard rigorous conceptualization and measurement of cultural
and production coordination effects in long-term buyer– differences. Journal of International Business Studies, 32(3):
supplier relationships. Academy of Management Journal, 519–535.
51(4): 785–807. Slangen, A. H. L., & van Tulder, R. J. M. 2009. Cultural distance,
Mesquita, L. F., & Lazzarini, S. G. 2008. Horizontal and vertical political risk, or governance quality? Towards a more accurate
relationships in developing economies: Implications for SMEs’ conceptualization and measurement of external uncertainty in
access to global markets. Academy of Management Journal, foreign entry mode research. International Business Review,
51(2): 359–380. 18(3): 276–291.
Meyer, K. E., & Nguyen, H. V. 2005. Foreign investment Taylor, C. R., Zou, S., & Osland, G. E. 1998. A transaction cost
strategies and sub-national institutions in emerging markets: perspective on foreign market entry strategies of US and
Evidence from Vietnam. Journal of Management Studies, 42(1): Japanese firms. Thunderbird International Business Review,
63–93. 40(4): 389–412.
Meyer, K. E., Estrin, S., Bhaumik, S. K., & Peng, M. W. 2009. Teece, D. J. 1986. Transactions cost economics and the
Institutions, resources, and entry strategies in emerging multinational enterprise: An assessment. Journal of Economic
economies. Strategic Management Journal, 30(1): 61–80. Behavior & Organization, 7(1): 21–45.
Nakos, G., & Brouthers, K. D. 2002. Entry mode choice of SMEs Van de Veijver, F., & Hambleton, R. K. 1996. Translating
in Central and Eastern Europe. Entrepreneurship Theory and tests: Some practical guidelines. European Psychologist, 1(2):
Practice, 27(1): 47–63. 89–99.
Norton, E. C., Wang, H., & Ai, C. 2004. Computing inter- Williams, D. E. 1992. Motives for retailer internationalization:
action effects and standard errors in logit and probit models. Their impact, structure and implications. Journal of Marketing
The Stata Journal, 4(2): 154–167. Management, 8(3): 269–285.
Oviatt, B. M., & McDougall, P. P. 1994. Toward a theory of Williamson, O. E. 1985. The economic institutions of capitalism:
international new ventures. Journal of International Business Firms, markets, relational contracting. New York: Free
Studies, 25(1): 45–64. Press.
Oxley, J. E. 1999. Institutional environment and the mechanisms Williamson, O. E. 1991. Comparative economic organization:
of governance: The impact of intellectual property protection The analysis of discrete structural alternatives. Administrative
on the structure of inter-firm alliances. Journal of Economic Science Quarterly, 36(2): 269–296.
Behavior & Organization, 38(3): 283–309. Williamson, O. E. 1998. Transaction cost economics: How it
Pan, Y., & Tse, D. K. 2000. The hierarchical model of market entry works; where it is headed. De Economist, 146(1): 23–58.
modes. Journal of International Business Studies, 31(4): 535–554. Zelner, B. A. 2009. Using simulation to interpret results from
Podsakoff, P. M., & Organ, D. W. 1986. Self-reports in logit, probit, and other nonlinear models. Strategic Manage-
organizational research: Problems and prospects. Journal of ment Journal, 30(12): 1335–1348.
Management, 12(4): 531–544. Zhao, H., Luo, Y., & Taewon, S. 2004. Transaction cost
Poppo, L., & Zenger, T. 2002. Do formal contracts and relational determinants and ownership-based entry mode choice:
governance function as substitutes or complements? Strategic A meta-analytical review. Journal of International Business
Management Journal, 23(8): 707–725. Studies, 35(6): 524–544.
Provan, K. G. 1993. Embeddedness, interdependence, and
opportunism in organizational supplier–buyer networks. Jour-
nal of Management, 19(4): 841–856. ABOUT THE AUTHORS
Puck, J. F., Holtbrügge, D., & Mohr, A. T. 2009. Beyond
entry mode choice: Explaining the conversion of joint Birger Maekelburger is Research Assistant and PhD
ventures into wholly owned subsidiaries in the People’s student at the chair of Business Administration,
Republic of China. Journal of International Business Studies,
40(3): 388–404. specializing in human resource management, small
Rasheed, H. S. 2005. Foreign entry mode and performance: The and medium-sized enterprises, and entrepreneur-
moderating effects of environment. Journal of Small Business ship at the University of Giessen. His research
Management, 43(1): 41–54.
Sapienza, H. J., Autio, E., George, G., & Zahra, S. A. interests include foreign market entry mode
2006. A capabilities perspective on the effects of early choice of SMEs, international entrepreneurship, and

Journal of International Business Studies

 Asset specificity and SME entry mode Birger Maekelburger et al
476

strategic orientation of young technology firms. He Ruediger Kabst is Professor of Business Administra-
was born in and is a citizen of Germany. tion, specializing in human resource management,
small and medium-sized enterprises, and entrepre-
Christian Schwens is Professor of Management at neurship, at the University of Giessen. His research
the Faculty of Business Administration and Econo- interests include international comparative human
mics of the Heinrich-Heine-University, Düsseldorf. resource management, expatriate management,
His research interests include foreign market entry interfirm cooperations (e.g., joint ventures),
mode choice of SMEs and international entrepre- international entrepreneurship, and entry mode
neurship. He was born in and is a citizen of choice of SMEs. He was born in and is a citizen
Germany. of Germany.

Accepted by Gabriel Szulanski, Area Editor, 20 January 2012. This paper has been with the authors for two revisions.

Journal of International Business Studies