FDI SPILLOVERS OVER TIME IN AN EMERGING MARKET: THE …haiyang/AMJ FDI spillover 2014.pdfTHE ROLES OF ENTRY TENURE AND BARRIERS TO IMITATION YAN ANTHEA ZHANG Rice University ... spillovers

FDI SPILLOVERS OVER TIME IN AN EMERGING MARKET:THE ROLES OF ENTRY TENURE AND BARRIERS

TO IMITATION

YAN “ANTHEA” ZHANGRice University

YU LIUniversity of International Business and Economics, China

HAIYANG LIRice University

In this study, we examine how foreign direct investment (FDI) spillovers to domesticfirms in an emerging market occur over time. From the organizational learningperspective, we propose that, as entry tenure of foreign firms in an industry increases,domestic firms can learn from the foreign firms over time and improve their produc-tivity. We further build upon the competitor imitation argument to propose that thiseffect will be stronger when barriers to imitation faced by the domestic firms are lower.Based upon a comprehensive panel dataset on manufacturing firms in China in1998–2007, our findings strongly support these arguments. We find that entry tenure offoreign firms in an industry has a positive relationship with the productivity ofindividual domestic firms in the same industry, albeit at a diminishing rate. We alsofind that this positive relationship is stronger when the foreign firms have lower exportintensity, lower intangible asset intensity, and have followed a more rhythmic (i.e., lessirregular) entry pattern—situations characterizing lower barriers to imitation.

Governments, especially those of emerging mar-ket countries such as China, have put great empha-sis on attracting foreign direct investment (FDI). Ithas been argued that, in addition to creating em-ployment opportunities and export income, foreign

firms from developed countries can create positivespillovers to domestic firms in emerging markets,which will be reflected in productivity increasesfor the latter (e.g., Blomström & Kokko, 1998; Buck-ley, Clegg, & Wang, 2007; Feinberg & Majumdar,2001; Spencer, 2008; Zhang, Li, Li, & Zhou, 2010).Most empirical studies on this topic, however, havetaken a snapshot approach to examine how thepresence of FDI (i.e., the ratio of FDI in an industry)affects domestic firms, and have produced incon-sistent findings. While some studies have foundpositive spillover effects (e.g., Buckley et al., 2007;Tian, 2007), others have demonstrated that foreignfirms have no spillover effect or even a negativeeffect (e.g., Aitken & Harrison, 1999; Feinberg &Majumdar, 2001).

To reconcile the mixed findings, recently, schol-ars have argued that FDI spillovers do not ariseautomatically but take place over time (De Backer &Sleuwaegen, 2003; Kosová, 2010; Spencer, 2008).Spencer (2008: 347), for example, noted that posi-tive FDI spillovers are likely to occur in the longterm because “benefits stemming from access to[foreign firms’] knowledge are likely to take a sub-

We would like to thank the editor, Dr. Jason Colquitt,and the three anonymous referees for their constructivecomments and suggestions that have helped significantlyimprove the paper. We have also benefited considerablyfrom the comments and suggestions of the participants ofthe 2011 Management Conference of China Europe Inter-national Business School (CEIBS) and the seminar par-ticipants at the Department of Management Science andEngineering of Stanford University and the École Poly-technique Fédérale de Lausanne (EPFL) School of Busi-ness, Switzerland. Yan “Anthea” Zhang appreciates thesupport of the National Natural Science Foundation ofChina (NSFC) (project number: 71232009) and the re-search support of CEIBS. Yu Li appreciates support fromthe NSFC (project number: 71002002) and the Funda-mental Research Funds for the Central Universities inUIBE (project number: 14JQ04). Haiyang Li appreciatessupport from the NSFC (project number: 71132007) andthe research support of CEIBS.

698

� Academy of Management Journal2014, Vol. 57, No. 3, 698–722.http://dx.doi.org/10.5465/amj.2011.0351

Copyright of the Academy of Management, all rights reserved. Contents may not be copied, emailed, posted to a listserv, or otherwise transmitted without the copyright holder’s expresswritten permission. Users may print, download, or email articles for individual use only.

stantial period of time to accrue” (emphasisadded). De Backer and Sleuwaegen (2003) foundthat new foreign entries decreased domestic firms’entry rate and increased their exit rate; however,the extant presence of foreign firms increased do-mestic firms’ entry rate and decreased their exitrate. Kosová (2010) found that initial foreign entryincreased the exit rate of domestic firms, but salesgrowth of the existing foreign firms increased thegrowth rate and survival rate of domestic firms.These findings indicate that there may exist crowd-ing-out effects on domestic firms upon foreign en-tries, but foreign firms may contribute positive ben-efits to domestic firms in the long run. Thus, it isimportant for FDI spillover research to shift fromthe snapshot toward a more dynamic approach.

Note that both De Backer and Sleuwaegen (2003)and Kosová (2010) focused on comparing the effectof new foreign entries with that of existing ones.Their findings thus shed few insights on how theeffect of FDI on domestic firms may occur over thecourse of the FDI presence. Moreover, from a theo-retical perspective, how FDI spillovers occur overtime may be a more complex process than thatwhich Spencer (2008) has argued. Different fromlearning in international joint ventures (IJVs), inwhich partners have incentives to teach each other(e.g., Inkpen & Tsang, 2005), FDI spillovers areunintended movements of knowledge from foreignfirms to domestic firms without compensation—or,a “free lunch” (Eden, 2009). Rather than teachingdomestic firms, foreign firms often take great effortsto protect their knowledge from being imitated bydomestic firms, especially in emerging marketswhere intellectual property rights are not well pro-tected (Feinberg & Gupta, 2009; Zhang, Li, Hitt, &Cui, 2007; Zhao, 2006). Thus, in the context of FDIspillovers, domestic firms’ learning from foreignfirms is what Nelson and Winter (1982: 124) wouldrefer to as “imitation from a distance.” This pointhas important implications for our understandingof how FDI spillovers may occur over time. First,since FDI spillovers are domestic firms’ “imitationfrom a distance,” there will be a threshold on howmuch they can learn over time. Second, how muchdomestic firms can learn from foreign firms overtime will depend upon the extent to which foreignfirms are imitable “from a distance.”

Our study is motivated by the desire to betterunderstand the dynamics of foreign firms’ spill-overs in an emerging market. We contribute to theliterature by explicitly examining the effect oftime—more specifically, the “entry tenure” of for-

eign firms—in FDI spillovers. An individual for-eign firm’s entry tenure in a host country refers tothe time elapsed since the firm was founded in thehost country (Zaheer & Mosakowski, 1997). Sincemultiple foreign firms may exist in an industry andthey may have entered at different time periods, forthe purpose of our study, we focus on the entrytenure of foreign firms at the industry level, definedas the average of all individual foreign firms’ entrytenure in an industry.1 Drawing upon the FDI spill-over literature and the organizational learning per-spective (Cyert & March, 1963; Kogut & Zander,1992; Winter, 2000), we propose that, as the entrytenure of foreign firms in an industry increases,domestic firms can improve their productivity bylearning advanced technologies and managementpractices from the foreign firms over time. How-ever, we do not expect this positive trend to belinear. Instead, domestic firms’ productivity willlikely increase at a diminishing rate as the foreignfirms’ entry tenure continues to increase, becauseobvious learning opportunities will gradually dryup. Our focus on foreign firms’ entry tenure allowsus to examine how FDI spillovers occur over thecourse of FDI presence, and, thus, provide directevidence on the dynamics of FDI spillovers.

Moreover, we draw upon the “competitor imita-tion” argument to examine how barriers to imita-tion may affect the effect of time in FDI spillovers.Reed and DeFillippi (1990: 89, emphasis added)argued that, “The ease with which competitive ad-vantage may be sustained or, alternatively, thespeed with which it is subject to imitation dependsupon the height of the barriers [to imitation].” Bar-riers to imitation can be created by informationasymmetry between a firm with competitive advan-tage and its competitors and/or causal ambiguity ofthe focal firm’s competitive advantage (Lippman &Rumelt, 1982; Reed & DeFillippi, 1990). We pro-pose that some important attributes of foreign firmscan affect the height of imitation barriers faced bydomestic firms: their market focus (export vs. do-

1 FDI can take the forms of wholly owned foreign firmsand foreign–domestic joint ventures. Domestic firmswith minority foreign ownership can benefit from knowl-edge flow from their foreign partners, in addition topossible spillovers from foreign firms with which theydo not have ownership relationships. To separate thespillover effect from the joint venture effect, we defined“foreign firms” as those with 100% foreign ownershipand defined “domestic firms” as those with 100% do-mestic ownership.

2014 699Zhang, Li, and Li

mestic market), asset composition (intangible as-sets vs. tangible assets), and the rhythm of theirentry pattern (a rhythmic vs. an irregular entrypattern). By shaping the height of the barriers toimitation, these attributes of foreign firms can affectthe role of time in FDI spillovers: the lower thebarriers to imitation, the stronger the positive rela-tionship between foreign firms’ entry tenure anddomestic firms’ productivity in an industry. Basedupon a comprehensive panel dataset of Chinesemanufacturing firms in 1998–2007, our findingssupport these arguments. Our findings hold with anumber of robustness checks, including using do-mestic firms’ survival as an alternative outcomevariable, using various definitions of domesticfirms and foreign firms, using various lags for pre-dictors, as well as using random subsamples withvarious sample sizes.

THEORY AND HYPOTHESIS DEVELOPMENT

FDI Spillovers Over Time

FDI spillovers in an emerging market represent aprocess in which domestic firms learn from foreignfirms (Meyer & Sinani, 2009; Zhang et al., 2010).Such learning occurs because, in an emerging mar-ket, foreign firms typically enjoy technological su-periority and strong management capabilities thatcan be transferred to or imitated by domestic firms(Sjöholm, 1999). Domestic firms, due to their tech-nology gaps with foreign firms, also have strongmotivation to learn from their foreign counterparts(Hitt, Li, & Worthington, 2005).

Learning, however, takes time to occur. Cyert andMarch (1963) addressed organizational learning asa process by which organizations as collectiveslearn through interactions with their environments.In this process, members of the organization searchfor and share information and knowledge from dif-ferent sources, creating organizational memory inthe form of shared beliefs, assumptions, and norms.These members need time to function as a sharedcognition system because “their work as learningagents is unfinished until the results of theirinquiry—their discoveries, inventions, and evalua-tions—are recorded in the media of organizationalmemory” (Argyris & Schön, 1978: 20). Because al-ternatives must be first created and an alternativecan be effectively assessed only by implementingit, the learning process is likely to be time consum-ing (Winter, 2000: 984). Thus, organizational learn-ing is a function of time (Simon, 1991).

Built upon this logic, we argue that FDI spillovers,in which domestic firms learn from foreign firms, willoccur as the entry tenure of foreign firms in an indus-try increases. The literature has identified four majorspillover mechanisms—demonstration effect, em-ployee turnover, domestic business linkages, andcompetitive pressure (Blomström & Kokko, 1998;Spencer, 2008)—all of which require a certainamount of time to effect. For demonstration effect, asthe entry tenure of foreign firms in an industry in-creases, domestic firms have a longer time to identify,imitate, and assimilate technologies and managementpractices used by the foreign firms. They also havemore time to try different combinations of theseknowledge components, assess the alternative com-binations, and utilize the best combination to createtheir competitive advantage.

Spillovers via employee turnover take time tooccur because foreign firms need to hire local em-ployees and train them before they can diffuse theforeign firms’ technologies and management prac-tices to their subsequent domestic employers. A2010 survey of human resources (HR) managersfrom 1,143 firms in China2 found that 27% of for-eign firms’ HR managers agreed that they facedcompetitive pressure from domestic non-state-owned firms in talent recruiting and retention,whereas only 17% of HR managers of domesticnon-state-owned firms agreed that they faced com-petitive pressure from foreign firms. Furthermore,according to this survey, domestic non-state-ownedfirms were attractive to job candidates, especiallythose at managerial level, because they providedbetter long-term career advancement prospects andmore attractive compensation packages.

Foreign firms’ local business linkages, particu-larly with local suppliers and distributors, are an-other important spillover mechanism. As foreignfirms transfer their technologies and know-how tolocal suppliers and distributors, such technologiesand know-how may be, ultimately, transmitted todomestic firms that use the same suppliers anddistributors (Spencer, 2008). However, building lo-cal business linkages is time consuming. In theearlier years of their presence, foreign firms oftenuse their parent firms’ existing (overseas) suppliersand distributors. As their entry tenure increases,

2 The survey was conducted by ManpowerGroup, aninternational recruitment firm. For further details, seehttp://www.manpower.com.cn/surveyreport.html (note:typo in original URL).

700 JuneAcademy of Management Journal

they gradually develop local business linkages(Belderbos, Capannelli, & Fukao, 2001; Spencer,2008) for cost reasons and/or because of the hostcountry’s government’s requirement (Osland &Björkman, 1998). In many emerging markets, in-cluding China, the host government’s “local com-ponents” requirement asks foreign firms to use acertain percentage of locally produced compo-nents, and this percentage typically increases asforeign firms’ entry tenure increases (Osland &Björkman, 1998).

The fourth mechanism, the competitive pressure,works as the increased competition that accompa-nies foreign entries puts pressure on domestic firmsand forces them to update their technologies andadopt advanced management practices to meet thiscompetitive pressure (Blomström & Kokko, 1998).While the effect of competitive pressure does notnecessarily require domestic firms’ learning fromforeign firms, in reality, competitive pressure typi-cally motivates domestic firms to learn from foreignfirms. The responding and technology upgradingprocesses also take time.

Most of the previous studies have investigatedforeign firms’ spillovers to domestic firms by esti-mating domestic firms’ productivity change (e.g.,Aitken & Harrison, 1999; Blomström & Kokko,1998; Buckley et al., 2007; Feinberg & Majumdar,2001; Zhang et al., 2010). As Görg and Strobl (2001:723) noted:

Since MNCs [multinational companies] use a higherlevel of technology, and technology, or knowledge,has certain characteristics of public goods (Caves,1996; Markusen, 1995), there is scope for technolog-ical externalities and indigenous firms may benefitthrough spillovers from MNCs. If there are produc-tivity spillovers, the presence of MNCs leads to pro-ductivity increases in domestic firms, allowingthem to become more efficient.

Following this tradition, we propose:

Hypothesis 1. The entry tenure of foreign firmsin an industry has a positive relationship withthe productivity of an individual domestic firmin the same industry.

Further, we expect that domestic firms’ produc-tivity will improve at a diminishing rate as foreignfirms’ entry tenure continues to increase. In theearly years of foreign firms’ presence, technologiesand management practices brought by these firmsare often very new to domestic firms, thus provid-ing ample opportunities for them to learn and im-

itate. As foreign firms’ entry tenure continues toincrease, obvious opportunities for domestic firmsto learn from the foreign firms—particularly, theircodified and discrete knowledge that may be trans-mitted by careful observation and without exten-sive interactions (Spencer, 2008)—gradually dryup. Foreign firms’ remaining competitive advan-tage is likely to be created by knowledge that is soidiosyncratic and tacit that “even successful repli-cation becomes problematic, let alone imitationfrom a distance” (Nelson & Winter, 1982: 124).Therefore, there tends to be a limit on the extent towhich domestic firms can improve their productiv-ity by learning from foreign firms over time. Thus:

Hypothesis 2. The positive relationship betweenthe entry tenure of foreign firms in an industryand the productivity of an individual domesticfirm in the industry becomes weaker as the for-eign firms’ entry tenure continues to increase.

Barriers to Imitation and the Role of Time in FDISpillovers

As noted above, FDI spillovers are different frominterorganizational learning in IJVs. In IJVs, mutualinterests and trust motivate partners to shareknowledge, and contracts and/or organizationalroutines guide how interorganizational learning oc-curs (e.g., Inkpen & Tsang, 2005; Steensma, Tih-anyi, Lyles, & Dhananraj, 2005). However, in FDIspillovers, foreign firms do not “teach” domesticfirms. Instead, they often take great effort to protecttheir knowledge from imitation by domestic firms(Feinberg & Gupta, 2009; Zhang et al., 2007; Zhao,2006). Thus, domestic firms’ learning from foreignfirms in FDI spillovers is what Nelson and Winter(1982: 124) refer to as “imitation from a distance.”

Prior studies have argued that the attributes ofdomestic firms—particularly, their ability to learnand motivation to learn—affect the extent to whichthey can benefit from FDI spillovers (e.g., Sjöholm,1999; Zhang et al., 2010). While this line of re-search provides important insights on FDI spill-overs, it only taps one side of the spillover relation-ship (i.e., domestic firms as the knowledge searchfirms), but ignores the other side of the relationship(i.e., foreign firms as the knowledge source firms).Since FDI spillovers are the outcome of domesticfirms’ imitation of foreign firms “from a distance,”the extent to which foreign firms are imitable “froma distance” should affect how much domestic firmscan learn from the foreign firms over time.


Barriers to imitation, or “the restraining or ob-structing of imitation by competitors,” can be cre-ated by information asymmetry between a firmwith competitive advantage and its competitorsand/or the causal ambiguity of the focal firm’s com-petitive advantage (Reed & DeFillippi, 1990: 94).Information asymmetry increases barriers to imita-tion because, if competitors cannot get sufficientinformation about a firm with competitive advan-tage, it would be hard for the competitors to imitatethe focal firm (Porter, 1985). Causal ambiguity re-fers to the “basic ambiguity concerning the natureof the causal connections between actions and re-sults” (Lippman & Rumelt, 1982: 420). It createsbarriers to imitation because, even if competitorshave full information about a firm, they may notunderstand how the firm’s competitive advantagearises, making it difficult for them to emulate thefocal firm’s strategy (Lippman & Rumelt, 1982;Reed & DeFillippi, 1990).

As Reed and DeFillippi (1990) argued, the speedwith which a firm’s competitive advantage is sub-ject to its competitors’ imitation depends upon thebarriers to imitation. The marginal effect of theentry tenure of foreign firms in an industry onthe productivity of individual domestic firms in theindustry can be viewed as the speed with which thedomestic firms learn from the foreign firms to im-prove their productivity. Following Reed and De-Fillippi’s (1990) argument, we propose that thepositive relationship between the entry tenure offoreign firms in an industry and domestic firms’productivity depends upon the barriers to imitationfaced by the domestic firms: the lower the barriers,the stronger the positive relationship. In this sec-tion, we investigate an array of foreign firms’ attri-butes—their market focus, asset composition, andrhythm of their entry pattern—that may affect theheight of barriers to imitation.

The moderating effect of foreign firms’ marketfocus. Market focus represents an important strate-gic choice of foreign firms operating in an emergingmarket (Luo & Park, 2001; Zhang et al., 2007).Those with an export market focus emphasize ex-ploiting the resource endowments of the host coun-try to meet demands of overseas markets. In con-trast, those with a local market focus emphasizereaping benefits from pent-up indigenous demandsin the host country (Luo & Park, 2001). We arguethat, relative to export focus, foreign firms’ localmarket focus can lower barriers to imitation facedby domestic firms. First, foreign firms’ local marketfocus decreases information asymmetry between for-

eign and domestic firms. As foreign firms offer prod-ucts to meet indigenous demands of the host countrymarket, spillovers via the demonstration effect be-come more likely to occur as domestic firms are ex-posed to foreign firms’ product offerings (Zhang et al.,2007). Such foreign firms also have strong incentivesto share their knowledge with local distributors andmarketing agents in order to penetrate the domesticmarkets, which increases the likelihood that theirtechnology and know-how will be transmitted to do-mestic competitors via their local business linkages(Zhang et al., 2007). In contrast, export-focused for-eign firms sell few or no products in the host countryand have few interactions with local distributors andmarketing agents. Since domestic firms do not havemuch information about these export-oriented for-eign firms, it is difficult for them to imitate the for-eign firms.

Second, local market-focused foreign firms havemore competitive interactions with domestic firms(Meyer & Sinani, 2009) than do export-focused ones.Direct competitive interactions with foreign firms inthe host country market motivate domestic firms toproactively search for information about the foreigncompetitors by closely following and back-engineer-ing their product offerings and accessing their localbusiness linkages. Domestic firms’ proactive informa-tion search will further reduce information asymme-try. As Reed and DeFillippi (1990) noted, aggressivecompetition can lead to faster decay in barriers toimitation. Relatedly, domestic firms in general have abetter understanding of the host country market thando foreign firms. To the extent that foreign firms focuson local markets, domestic firms’ local knowledgeenables them to better understand foreign firms’ ac-tions and their performance implications (i.e., whatwill work and what won’t), which can facilitate theirlearning from the foreign firms. Domestic firms caneven combine their local knowledge and new knowl-edge elements brought by foreign firms to becomebetter than foreign firms in the local markets. Com-bining these arguments together, we propose that:

Hypothesis 3. The positive relationship be-tween the entry tenure of foreign firms in anindustry and the productivity of an individualdomestic firm in the industry is stronger whenthe foreign firms have lower export intensity.

The moderating effect of foreign firms’ assetcomposition. Firm assets can be generally dividedinto two broad categories: tangible assets, such asland, plants, equipment, inventories, and otherphysical assets; and intangible assets, such as pat-


ents, trademarks, copyrights, and trade secrets.Barth, Kasznik, and McNichols (2001) argued thatfirms with substantial intangible assets have greaterinformation asymmetry between insiders and out-siders. Although the value of intangible assets maybe estimated and reported in firms’ financial dis-closures, their substance is less observable to out-siders. Also, firms typically purchase tangible assetsmore frequently than intangible assets (Barth et al.,2001; Teece, 1998). Thus, firms’ tangible assets aremore likely to be observed and imitated by competi-tors via market transactions. Competitors can evenpurchase the same or similar tangible assets from thefocal firms’ suppliers. For these reasons, we arguethat foreign firms with higher intangible asset inten-sity have greater information asymmetry betweenthem and their domestic competitors, which in-creases the barriers to imitation. Similarly, Tian(2007) argued that tangible assets, such as equipmentand production lines, are hard to be protected fromdomestic firms’ imitation whereas intangible assets,such as patents and secret formulae and ingredients,are normally better protected from domestic firms’“stealing.” Moreover, Meyer and Sinani (2009: 1077)argued that foreign firms with a higher level of intan-gible assets are likely to operate in the more upmarketsegments in an emerging market, in which “they ex-perience little direct interaction with local firms op-erating in volume-driven mass markets with smallmargins.” Different market focuses can further in-crease information asymmetry between foreign anddomestic firms.

Intangible assets also have more uncertain pay-offs than tangible assets (Barth et al., 2001). This isbecause payoffs of intangible assets tend to be firm-specific and depend upon what other (tangible andintangible) assets (i.e., complementary assets) thefirm has (Teece, 1998). Thus, for foreign firms withhigher intangible asset intensity, the causal rela-tionship between their actions and performance isless transparent to domestic firms because of theuncertain payoffs of their intangible assets. There-fore, intangible assets of foreign firms can also cre-ate higher barriers of imitation by increasing casualambiguity of their competitive advantage. For thesereasons, we propose that:

Hypothesis 4. The positive relationship be-tween the entry tenure of foreign firms in anindustry and the productivity of an individualdomestic firm in the industry is stronger whenthe foreign firms have lower intangible assetintensity.

The moderating effect of foreign firms’ entrypattern. The pattern with which foreign firms enteran industry in the host country may also have im-portant implications for barriers to imitation. Intheir study on firms’ internationalization process,Vermeulen and Barkema (2002) investigated howthe pattern of a firm’s foreign expansion may affectthe extent to which the firm can learn from its owninternationalization experience. They argued thatfirms following a more rhythmic process can learnbetter from their experience because they can relatetheir current foreign expansions to their recentpast. In contrast, firms following a more irregularprocess tend to overstretch their absorptive capac-ity at peaks of rapid foreign expansions and forgettheir experience in the long periods of inactivity.Empirically, they found that the positive relation-ship between the number of a firm’s foreign sub-sidiaries and its performance was stronger if thefirm had followed a more rhythmic international-ization process. Different from Vermeulen andBarkema’s (2002) work, which focused on the pat-tern of a firm’s own foreign expansions (i.e., within-firm pattern) and investigated how this pattern mayaffect the firm’s learning from its own experience,we focus on the pattern of foreign firms’ entries(i.e., across-firm pattern) at the industry level andinvestigate how this pattern can affect domesticfirms’ learning from the foreign firms (other firms).

Based upon Vermeulen and Barkema (2002), weconstructed two figures (Appendix A) that depicttwo different entry patterns. In a “rhythmic entrypattern,” foreign firms enter an industry in a con-stant and stable pace—for example, a couple offoreign entrants every year. In contrast, in an “ir-regular pattern,” foreign entries are concentrated ina few time periods, followed by long periods withno foreign entries. We argue that, relative to anirregular entry pattern, a rhythmic entry pattern offoreign firms can lower barriers to imitation facedby domestic firms. In an irregular pattern, a largenumber of foreign firms enter the industry together.It is overwhelming for domestic firms to search for,interpret, and understand technologies and prac-tices brought by the foreign entrants. Their absorp-tive capability can be overstretched, and theymay not be able to devote sufficient time and effortto each of the new entrants. As a result, some im-portant aspects of the foreign entrants may not benoticed or understood, creating greater causal am-biguity regarding the foreign firms’ competitive ad-vantage. Moreover, according to the “time compres-sion diseconomies” argument (Dierickx & Cool,


1989), experiences that come too fast will generatefewer benefits than the same experiences wouldhave if they had come in a more gradual, absorbablepattern. So, if foreign firms have followed an irreg-ular entry pattern, domestic firms’ learning benefitsmay be limited due to the time compression disec-onomies.3

Conversely, if foreign firms have followed arhythmic entry pattern, domestic firms only needto pay attention to and observe a few new entrantsat a time. The focused attention enables the domes-tic firms to observe various aspects of the newforeign entrant(s)’ operations and thus get a morecomprehensive understanding of their competitiveadvantage. Moreover, domestic firms can graduallydevelop their knowledge stock by learning steadilyfrom earlier foreign entrants. Thus, they can havestronger absorptive capacity (Cohen & Levinthal,1990) to draw upon the knowledge elementsbrought by subsequent foreign entrants. As Brownand Eisenhardt (1997: 24) found in the computerindustry, when firms implement and absorbchanges in a rhythmical pattern, they may create aflow of attention and “become focused, efficient,and even more confident about the task at hand.”For the above reasons, we propose that:

Hypothesis 5. The positive relationship be-tween the entry tenure of foreign firms in anindustry and the productivity of an individualdomestic firm in the industry is stronger if theforeign firms have followed a more rhythmic(i.e., less irregular) pattern of entering theindustry.

METHODOLOGY

Data Sources and Sample

We tested these hypotheses in the context ofChina’s emerging market, the largest FDI recipientcountry in the world. The major data source for thisstudy was the Annual Industrial Survey Database(1998–2007) from the Chinese National Bureau ofStatistics (CNBS). The database contained the mostcomprehensive information about domestic andforeign firms in China (Chang & Xu, 2008; Tian,2007). By law, all firms in China are required tocooperate with and submit their basic and financialinformation to the CNBS (Chang & Xu, 2008). Thedatabase included key firm-level financial informa-tion such as sales, capital, and employment, as wellas demographic information such as the foundingyear and ownership details. The aggregation of thefirm-level information is published in the officialChina Statistical Yearbook. CNBS statistics havebeen used by previous studies in the strategy andinternational business areas (e.g., Chang & Xu,2008; Tian, 2007; Zhang et al., 2010).

Since 1998, the database covered all state-ownedfirms, and non-state-owned firms (including for-eign ones) with annual sales of RMB 5 million(about US$670,000 according to the official ex-change rate at the end of 2007) or above. Hence,data prior to 1998 were not included in the study toensure consistency in firm coverage. Also, consis-tent with previous studies on FDI spillovers (e.g.,Javorcik, 2004; Tian, 2007; Zhang et al., 2010), wefocused on firms in the manufacturing industries.As noted in Footnote 1, to separate spillover effectsfrom joint venture effects, we defined “foreignfirms” as firms with 100% foreign ownership and“domestic firms” as firms with 100% domesticownership. Our sample consisted of 945,553 do-mestic firm–year observations, covering 301,667domestic firms (unevenly distributed across years)in 1998–2007.4 These domestic firms covered 511four-digit Standard Industrial Classification (SIC)code manufacturing industries, accounting formore than 93% of manufacturing industries in

3 One may argue that, in an irregular entry pattern, alarge number of foreign firms entered an industry in thehost country in a short time period; consequently, do-mestic firms would have little incentive to enter thisindustry since there are already a large number of playersin the market. To rule out this alternative explanation,we conducted additional analyses to investigate how theaccumulated number of foreign firms in an industry atthe end of the prior year (Yeart�1) (i.e., the stock offoreign firms) and the number of new foreign entries inthe industry in the prior year (Yeart�1) (i.e., the flow offoreign firms) may affect domestic firms’ entries in theindustry in a year (Yeart). Our results demonstrated thatthe stock of foreign firms encouraged subsequent domes-tic entries, while the flow of foreign firms had no signif-icant effect on subsequent domestic entries. Therefore,we are confident that our reported results were notdriven by this alternative explanation.

4 Data used in this study have some overlap with thedata used in Zhang et al. (2010). Zhang et al. (2010) usedCNBS data from 1998–2003, whereas this study usedCNBS data from 1998–2007. While these two studiesunavoidably had overlaps in control variables (e.g., shareof foreign firms in an industry), the independent variableand the three moderating variables used in this studywere not used in Zhang et al. (2010).


China. Further, data on 223,382 foreign firm–yearobservations from 64,946 foreign firms were usedto calculate FDI-related variables.

Model Specification

Consist with previous spillover studies (e.g., Ai-tken & Harrison, 1999; Tian, 2007; Zhang et al.,2010), we estimated how a domestic firm’s produc-tivity may be affected by foreign firms. It can beexpressed by a log-linear Cobb–Douglas productionfunction as follows.

Yijt� �1 LogKijt � �2 LogLijt � �3 Entry tenure

of foreign firms in an industryj(t�1)

� �4 Export intensity of foreign firms

in the industryj(t�1)

� �5 Intangible asset intensity of

foreign firms in the industryj(t�1)

� �6 Rhythm of entry pattern of

foreign firms in the industryj(t�1)

� �7 Squared term of entry tenure of

foreign firmsj(t�1)

� �8 Entry tenure of foreign firmsj(t�1)

� Export intensity of foreign firmsj(t�1)


� Intangible asset intensity of



� Rhythm of entry pattern of


� �11 Controls � �ij � �ijt (1)

Log output Yijt for domestic firm i in industrysector j (defined at the four-digit SIC code level) attime t was regressed on its inputs (Log Kijt and LogLijt) and the entry tenure of foreign firms in theindustry sector j at time t � 1 and the foreign firms’export intensity, intangible asset intensity, andrhythm of their entry pattern in sector j at time t �1, as well as their interaction terms. We used firmannual sales for firm output (Chung, Mitchell, &Yeung, 2003). To remove the effects of deflation orinflation due to price change over time, we deflatedfirm sales using the year 2000’s constant price (Ai-tken & Harrison, 1999; Tian, 2007). Firm inputsincluded the log of the firm’s capital input (Kijt, itscapital stock) and the log of its labor input (Lijt, its

number of employees). We deflated the capitalstock by the GDP deflator on the basis of the year2000’s constant price (Aitken & Harrison, 1999;Tian, 2007). A vector of controls (measured at yeart � 1) was also included. Further, �ij was an unob-served effect for domestic firm i in sector j, and ijt

was the error term. This model specification usedone-year lag for predictors and controls.

Measurement

Measurement of predictors. In this study, wefirst calculated the entry tenure of each foreignfirm5 in the industry, referring to the number ofyears since the firm was founded in China. We thencalculated the mean of individual foreign firms’entry tenure, weighted by their asset sizes, to mea-sure entry tenure of foreign firms at the industrylevel. We used the weighted measure to take intoaccount the fact that large foreign firms tend tohave greater impact on domestic firms than dosmall ones.

There are alternative ways to operationalize en-try tenure of foreign firms at the industry level,such as tenure of the earliest entrant and tenure ofthe largest entrant. However, we believe that ourmeasure is better than the other options. It is com-mon that small foreign investments were tried in anemerging market before larger commitments weremade (by the same foreign investors or by otherforeign investors). The measure of “tenure of theearliest entrant” would ignore the effects of later(probably larger and more important) foreign en-trants. On the other hand, the measure of “tenure ofthe largest entrant” would ignore foreign entrantsthat entered the host country before the largestentrant. Compared to these two alternatives, ourmeasure took into account the effects of all foreignentrants and their sizes.

In our data, the entry tenure of individual foreignfirms ranged between 1 and 30 years; the (mean of)entry tenure of foreign firms at the industry levelranged between 1 and 24 years. The wide ranges ofthese variables allowed us to test our hypotheses.We also used alternative foreign firm size measures(i.e., employment and sales size) as the weight, as

5 If a foreign investor had registered multiple subsid-iaries in China, each subsidiary was treated as a separateforeign firm. That is, our definition of foreign firms wasat the subsidiary level instead of at the corporate level.


well as simple average entry tenure of foreign firms,and got highly consistent results.

Export intensity of foreign firms in an industry wasmeasured as the sum of foreign firms’ export salesdivided by the sum of their total sales. Intangibleasset intensity of foreign firms in an industry wasmeasured as the sum of foreign firms’ intangible as-sets divided by the sum of their total assets. In theCNBS survey database, tangible assets included land,plants, equipment, inventories, and other physicalassets; intangible assets included patents, non-pat-ented technologies, trademarks, copyrights, trade se-crets, and other nonphysical assets (Tian, 2007).

We followed Vermeulen and Barkema’s (2002:644) measure of the rhythm of a firm’s internation-alization process to create the measure of rhythm offoreign firms’ entry pattern in an industry. In Ap-pendix A, the upper graphs depict the total numberof foreign firms in an industry by year and thebottom graphs depict the change in the number offoreign firms in an industry by year. Rhythm offoreign firms’ entry pattern was measured by thekurtosis of this distribution:

kurtosis � � n(n � 1)(n � 1)(n � 2)(n � 3)��xi � x

�

s�4�

�3(n � 1)2

(n � 2)(n � 3) (2)

where n � number of foreign firms in the industry,xi � entry tenure of foreign firm i, and s � standarddeviation of the entry tenure of foreign firms in theindustry. Values of the kurtosis in our data rangedbetween –6 and 11, with higher values indicatingmore irregular entry patterns and lower values in-dicating more rhythmic entry patterns. For ease ofinterpretation, we recoded it by multiplying (–1) sothat higher values indicate more rhythmic entrypatterns. All of these predictors were calculated foreach four-digit SIC code industry in the sample andupdated yearly.

Measurement of controls. Most previous studieson FDI spillovers have focused on the effect of FDIshare in an industry (e.g., Aitken & Harrison, 1999;Javorcik, 2004; Tian, 2007). We controlled for shareof foreign firms in an industry, measured by theratio of the foreign firms’ sales to the industry’stotal sales. Or, we measured it as the ratio of theforeign firms’ employment (or assets) to the indus-try’s total employment (or assets). These alternativemeasures produced virtually the same results. Inmany industries in China, there have been a mix of

wholly owned foreign firms and IJVs. To rule outthe possible effect of IJVs on domestic firms thatwere not the domestic partners of the IJVs, wecontrolled for the (logged) number of IJVs in theindustry in the prior year, referring to firms whereforeign ownership was equal to or larger than 50%but less than 100%. Since foreign partners are morelikely to contribute their technologies and skills toIJVs if they have majority ownership (e.g., Yan &Gray, 1994), IJVs in which foreign partners own50% or more should be more relevant from a spill-over perspective. Alternatively, we used 25% for-eign ownership as a cut-off (minimum foreign own-ership for being legally considered as a foreign firmin China) and got virtually the same results.

We also controlled for the following importantattributes of individual domestic firms that mayaffect their productivity and/or their motivation/ability to learn from foreign firms. A domesticfirm’s age was measured by the number of yearssince it was founded. Its squared term was alsocontrolled for. State ownership of a domestic firmwas coded “1” if a domestic firm was state ownedand “0” otherwise. Since some domestic firms’state ownership changed in our research time pe-riod, this variable was updated yearly. A domesticfirm’s export intensity may affect its productivityand its motivation/ability to learn from foreignfirms. It was measured as the ratio of a domesticfirm’s export sales to its total sales. A domesticfirm’s intangible asset intensity was also controlledfor and was measured as the ratio of the firm’sintangible assets to its total assets. A domesticfirm’s spatial relationships with foreign firms mayalso matter. We controlled for a domestic firm’scolocation density of foreign firms, measured bythe (logged) number of foreign firms in the industrythat were located in the domestic firm’s province.All domestic firm-related variables were updatedyearly. Moreover, to capture the possible effect ofChina’s economic change over time, we includedcalendar year dummies. We also controlled for re-gion dummies at the provincial level and industrydummies at the two-digit SIC code level.

Correction for Survival Bias

Our sample was an unbalanced panel of domes-tic firms with varying survival lengths. Firms withhigher productivity in one year are more likely tosurvive in the next year. Therefore, our estimationof individual domestic firms’ productivity was sub-ject to their survival bias. To correct for this possi-


ble bias, we estimated the likelihood that a domes-tic firm will survive in year t � 1, using predictorsincluding the domestic firm’s age and its squaredterm, its total profit, the average labor productivityof firms in the industry, and the number of foreignfirms in the industry and its squared term (all mea-sured in year t). The domestic firm’s size was notincluded because it was highly correlated with itstotal profit. Following Chatterjee and Hambrick(2007), we used the “xtprobit” command in thestatistical software package Stata for this estimationand calculated the inverse Mill’s ratio from theestimation (Data in year 2008 was added to calcu-late the inverse Mill’s ratio for firm observations inthe year 2007). The inverse Mill’s ratio was thenincluded as a control in our models estimatingindividual domestic firms’ production function.

Data Analyses

Based upon Wooldridge (2002), we adopted thefollowing procedures for data analyses. First, weused the Breusch-Pagan Lagrange multiplier test todecide whether the panel data method or thepooled ordinary least squares approach was moreappropriate. The results of the test suggested thatunobserved individual effects existed in the data.Thus, the panel data method was used because thismethod can model the unobserved individual ef-fects associated with the same units. Second, forthe panel data method, we needed to choose be-tween fixed effect and random effect models. Theresults of Hausman test revealed that explanatoryvariables were correlated with the unobserved ef-fects, and, thus, fixed effect models should be used.We therefore followed Wooldridge (2002: 267) andconducted a fixed effect transformation (also calledthe “within transformation”) to eliminate unob-servable firm effects. We added year dummies, in-dustry dummies, and region dummies after thefirm-fixed effect transformation in order to controlfor time-varying unobservable region and industryeffects (the firm-fixed effect transformation onlyeliminates the unobserved effects that do notchange over time), which may drive changes inattractiveness of a particular region or industry (Ja-vorcik, 2004: 616). Finally, we used the modifiedWald test to check for heteroscedasticity, and theresults suggested that this was an issue. We there-fore clustered standard errors at the firm level toadjust for heteroscedasticity.

RESULTS

Table 1 reports the descriptive statistics and cor-relations of the variables used in this study, exceptfor year dummies, region dummies, and industrydummies (for space reasons). Table 2 presents firm-fixed effect models of individual domestic firms’ pro-duction function. Model 1 features the controls only,Model 2 adds the predictors’ main effects, andModel 3 adds the interaction terms. Alternatively, weadded the interaction terms one by one and the re-sults were virtually the same as those for Model 3.Variables were mean-centered prior to creating theinteraction terms in order to reduce concerns of mul-ticollinearity (Aiken & West, 1991).

The coefficient of the entry tenure of foreignfirms is positive and significant (b � 0.016, p �0.001, Model 3 in Table 2), meaning that it has asignificantly positive relationship with a domesticfirm’s productivity. Hypothesis 1 is thus sup-ported. The coefficient for its squared term is neg-ative and significant (b � �0.002, p � 0.001,Model 3 in Table 2). To facilitate interpretation, weplotted this curvilinear effect in Figure 1, followingthe procedure recommended by Aiken and West(1991). As shown in Figure 1, the entry tenure offoreign firms in an industry has an overall positiverelationship with a domestic firm’s productivity, al-though the slope of this positive relationship dimin-ishes (but doesn’t turn negative) as the entry tenure ofthe foreign firms continues to increase. Hypothesis 2is thus supported. To demonstrate practical implica-tions of these results, we calculated the effect sizefollowing the approach used by Aitken and Harrison(1999: 609). All else being equal, as the entry tenure offoreign firms in an industry increased from 1 year to10 years, a domestic firm’s output would increase by14.4%.6 Since we have controlled for differences inthe domestic firm’s capital and labor input, the 14.4%increment is a pure total factor productivity gain,indicating significant practical importance.

6 As the entry tenure of foreign firms in an industryincreased from 1 year to 10 years, this variable had thefollowing vector of values: [1, 2, . . . , 9, 10]. The groupmean of this vector is 5.5. After fixed effect transfor-mation (minus the group mean of 5.5), the vector wastransformed to [– 4.5, –3.5, . . . , 3.5, 4.5]. All else beingequal, the difference in domestic firms’ productivity asthe entry tenure of foreign firms in an industry in-creased from 1 year to 10 years would be: {[0.016 � 4.5� (�0.002) � (4.5)2] � [0.016 � (�4.5) � (�0.002) �(�4.5)2]} � 0.144.


TA

BL

E1

Mea

ns,

Sta

nd

ard

Dev

iati

ons,

and

Cor

rela

tion

sof

Var

iabl

esa

,b,c

,d

Var

iabl

esM

SD

12

34

56

78

910

1112

13

1.L

ogY

9.74

01.

387

1.00

02.

Log

K9.

522

1.39

30.

659

1.00

03.

Log

L4.

738

1.09

50.

574

0.68

81.

000

4.E

ntr

yte

nu

reof

fore

ign

firm

sin

anin

du

stry

6.59

01.

861

0.06

3�

0.08

5�

0.10

51.

000

5.E

xpor

tin

ten

sity

offo

reig

nfi

rms

0.38

90.

257

�0.

064

�0.

150

0.01

60.

122

1.00

06.

Inta

ngi

ble

asse

tin

ten

sity

offo

reig

nfi

rms

0.03

40.

055

�0.

063

0.02

70.

023

�0.

194

�0.

159

1.00

07.

Rh

yth

mof

entr

yp

atte

rnof

fore

ign

firm

s�

0.11

81.

543

0.05

6�

0.01

2�

0.02

30.

129

0.07

6�

0.02

31.

000

8.S

har

eof

fore

ign

firm

s0.

140

0.13

0�

0.02

7�

0.12

1�

0.05

80.

269

0.50

7�

0.17

20.

134

1.00

09.

Nu

mbe

rof

IJV

s3.

293

1.26

50.

045

�0.

042

0.00

60.

207

0.15

4�

0.14

00.

008

0.26

41.

000

10.

Dom

esti

cfi

rm’s

age

11.1

0312

.79

�0.

067

0.21

50.

268

�0.

131

�0.

056

0.09

6�

0.08

5�

0.12

4�

0.11

41.

000

11.

Dom

esti

cfi

rm’s

stat

eow

ner

ship

0.16

30.

370

�0.

214

0.15

70.

137

�0.

190

�0.

145

0.15

4�

0.11

0�

0.17

0�

0.13

60.

457

1.00

012

.D

omes

tic

firm

’sex

por

tin

ten

sity

0.10

50.

285

0.04

2�

0.03

60.

115

0.08

20.

311

�0.

057

0.05

10.

215

0.11

9�

0.04

2�

0.10

01.

000

13.

Dom

esti

cfi

rm’s

inta

ngi

ble

asse

tsin

ten

sity

0.02

00.

060

0.04

00.

137

0.07

30.

021

�0.

030

0.01

50.

019

0.00

2�

0.00

30.

030

�0.

010

0.01

51.

000

14.

Dom

esti

cfi

rm’s

colo

cati

ond

ensi

tyof

fore

ign

firm

s1.

551

1.55

70.

088

�0.

106

�0.

064

0.23

60.

311

�0.

133

0.13

00.

432

0.56

1�

0.20

1�

0.26

80.

243

�0.

020

an

�94

5,55

3fi

rm–y

ear

obse

rvat

ion

s.b

Yea

rd

um

mie

s,re

gion

du

mm

ies,

and

ind

ust

ryd

um

mie

sar

en

otin

clu

ded

inth

ista

ble

for

spac

ere

ason

s,bu

tw

ere

incl

ud

edin

all

mod

els.

cC

orre

lati

ons

wit

hab

solu

teva

lues

equ

alto

orgr

eate

rth

an0.

002

are

sign

ific

ant

atth

ele

vel

ofp

�0.

05.

dC

orre

lati

onco

effi

cien

tsbe

twee

nso

me

con

trol

vari

able

sar

egr

eate

rth

an0.

4,w

hic

hm

ayra

ise

con

cern

sof

mu

ltic

olli

nea

rity

.In

sup

ple

men

tary

anal

yses

,w

ed

rop

ped

thes

eco

ntr

olva

riab

les

inm

odel

san

dob

tain

edh

igh

lyco

nsi

sten

tre

sult

s.

In Model 3 in Table 2, the interaction term of theentry tenure of foreign firms and the foreign firms’export intensity is negative and significant (b ��0.006, p � 0.001, Model 3 in Table 2). We plottedthis interaction effect, following the procedure rec-ommended by Aiken and West (1991), in Figure 2.The plot depicts that the positive relationship be-tween the entry tenure of foreign firms in an indus-try and the productivity of a domestic firm in theindustry is stronger when the foreign firms’ export

intensity is low (one standard deviation belowmean) than when it is high (one standard deviationabove mean). Thus, Hypothesis 3 is supported. Theinteraction term of the entry tenure of foreign firmsand the foreign firms’ intangible asset intensity isnegative and significant (b � �0.032, p � 0.001,Model 3 in Table 2). The plot of this interactioneffect in Figure 3 demonstrates that the positiverelationship between the entry tenure of foreignfirms in an industry and the productivity of a do-

TABLE 2Firm-Fixed Effect Models of Domestic Firm Production Functiona,b,c

Variables Model 1 Model 2 Model 3

Log K 0.382*** (0.002) 0.379*** (0.002) 0.379*** (0.002)Log L 0.376*** (0.003) 0.380*** (0.003) 0.380*** (0.003)PredictorsEntry tenure of foreign firms in an industry 0.016*** (4.73E-04) 0.016*** (4.74E-04)Export intensity of foreign firms –0.049*** (0.005) –0.051*** (0.005)Intangible asset intensity of foreign firms –0.054*** (0.011) –0.029* (0.011)Rhythm of entry pattern of foreign firms 0.005*** (4.13E-04) 0.005*** (4.13E-04)InteractionsSquared term of entry tenure of foreign

firms in an industry–0.002*** (1.16E-04)

Entry tenure of foreign firms in anindustry � Export intensity offoreign firms

–0.006*** (1.51E-03)

Entry tenure of foreign firms in anindustry � Intangible asset intensity offoreign firms

–0.032*** (0.004)

Entry tenure of foreign firms in an industry �Rhythm of entry pattern offoreign firms

6.95E-04*** (2.17E-04)

ControlsShare of foreign firms 0.045*** (0.011) 0.010 (0.011) 0.006 (0.011)Number of IJVs 3.63E-03** (1.24E-03) 1.78E-03 (1.25E-03) �2.12E-04 (1.26E-03)Domestic firm’s age 0.002*** (1.24E-04) 0.002*** (1.24E-04) 0.002*** (1.24E-04)Domestic firm’s age squared �5.88E-05*** (4.32E-06) �5.13E-05*** (4.27E-06) �5.11E-05*** (4.25E-06)Domestic firm’s state ownership 0.002 (0.002) 8.94E-04 (0.002) 0.001 (0.002)Domestic firm’s export intensity 0.029*** (0.004) 0.029*** (0.004) 0.029*** (0.004)Domestic firm’s intangible asset intensity –0.092*** (0.013) –0.101*** (0.013) –0.101*** (0.013)Domestic firm’s colocation density of foreign firms 0.026*** (0.001) 0.022*** (0.001) 0.021*** (0.001)Inverse Mill’s ratio –1.359*** (0.013) –1.287*** (0.013) –1.306*** (0.013)Region dummies Included Included IncludedIndustry dummies Included Included IncludedYear dummies Included Included IncludedF value 2,139*** 2,069*** 1,983***Within R2 0.2636 0.2651 0.2655Total R2 0.9306 0.9308 0.9309F value for R2 change — M2 vs. M1 241.37*** M3 vs. M2 74.07***

a n � 945,553 firm–year observations.b Estimated coefficients and associated robust standard errors (in parentheses) are reported.c The models do not have constants because the fixed effect transformation has removed the intercept term.

* p � .05 (two-tailed tests)** p � .01

*** p � .001


FIGURE 1Entry Tenure of Foreign Firms in an Industry and Domestic Firm Productivity: A Nonlinear Effect

FIGURE 2Entry Tenure of Foreign Firms in an Industry and Domestic Firm Productivity: The Moderating Role of

the Export Intensity of Foreign Firms


mestic firm in the industry is stronger when theforeign firms’ intangible asset intensity is low thanwhen it is high. Thus, Hypothesis 4 is also sup-ported. The interaction term of the entry tenure of

foreign firms and the rhythm of their entry patternis positive and significant (b � 6.95E-04, p �0.001). The plot of this interaction effect in Figure 4depicts that the positive relationship between the


the Intangible Assets Intensity of Foreign Firms


the Rhythm of Entry Pattern of Foreign Firms


entry tenure of foreign firms in an industry and theproductivity of a domestic firm in the industry isstronger if the foreign firms have followed a rhyth-mic entry pattern than if they have followed anirregular entry pattern. These results support Hy-pothesis 5.

Supplementary Analysis I: Domestic Firms’Survival as an Alternative Outcome Variable

In addition to domestic firms’ productivity, an-other important outcome variable in the FDI spill-over literature is the survival/exit of domestic firms(e.g., Chang & Xu, 2008; De Backer & Sleuwaegen,2003; Kosová, 2010). We used this alternative out-come variable to check the robustness of our find-ings. Exit of a domestic firm in a year was coded“1” if the domestic firm was included in the CNBSAnnual Industrial Survey Database in the prior yearbut not in the current year, and “0” otherwise. Weestimated the likelihood of a domestic firm’s exitusing the Gompertz model, following previousstudies on organizational mortality (e.g., Carroll& Delacroix, 1982; Freeman, Carroll, & Hannan,1983). A negative coefficient would mean that a

predictor has a negative (positive) relationshipwith the likelihood of domestic firms’ exit (sur-vival). Results are reported in Table 3. The entrytenure of foreign firms in an industry (b � �0.012,p � 0.001, Model 3 in Table 3) has a negative(positive) relationship with the likelihood of do-mestic firms’ exit (survival). The coefficient for itssquared term (b � 0.005, p � 0.001) is positive andsignificant, suggesting that the negative (positive)relationship between the entry tenure of foreignfirms in an industry and the likelihood of exit (sur-vival) of a domestic firm in the industry becomesweaker as the entry tenure of the foreign firmscontinues to increase. These results are consistentwith the predictions of Hypotheses 1 and 2, had adomestic firm’s survival been used as the depen-dent variable.

The interaction of the entry tenure of foreignfirms in an industry with the foreign firms’ exportintensity (b � 0.048, p � 0.001) and its interactionwith the foreign firms’ intangible asset intensity(b � 0.105, p � 0.01) are both positive and signif-icant. These results indicate that the negative (pos-itive) relationship between the entry tenure of for-

TABLE 3Models of Domestic Firm Exita,b,c


PredictorsEntry tenure of foreign firms in an industry –0.013*** (0.002) –0.012*** (0.002)Export intensity of foreign firms 0.076*** (0.016) 0.091*** (0.017)Intangible asset intensity of foreign firms –0.630*** (0.076) –0.801*** (0.080)Rhythm of entry pattern of foreign firms –0.006*** (1.75E-03) –0.005** (1.75E-03)InteractionsSquared term of entry tenure of foreign

firms in an industry0.005*** (4.94E-04)

Entry tenure of foreign firms in an industry �Export intensity of foreign firms

0.048*** (0.006)

Entry tenure of foreign firms in an industry �Intangible asset intensity of foreign firms

0.105** (0.036)

Entry tenure of foreign firms in an industry �Rhythm of entry pattern of foreign firms

0.001 (0.001)

Controls(All controls listed in Table 2 except inverse

Mill’s ratio and Log Kc were included.)Pseudo-log-likelihood �221,804 �221,747 �221,670Wald chi-square 5.54E�07*** 5.54E�07*** 5.49E�07***

a Number of firm–year observations � 945,553; number of exits � 113,633.b Estimated coefficients and associated robust standard errors (in parentheses) are reported.c Only one measure of firm size (Log L or Log K) was needed for inclusion in the model. We included Log L, but including Log K

produced virtually the same results.* p � .05 (two-tailed tests)

** p � .01*** p � .001


eign firms in an industry and the likelihood of exit(survival) of a domestic firm in the industry isstronger when the foreign firms’ export intensity islower, and/or when their intangible asset intensityis lower. These findings are consistent with thepredictions of Hypotheses 3 and 4. However, theinteraction of the entry tenure of foreign firms in anindustry and the rhythm of their entry patternis not significant. Overall, our results on domesticfirm survival are generally consistent with those ondomestic firm productivity, with the exception ofHypothesis 5.

Supplementary Analysis II: Addressing the Issueof Left-Censoring

In Table 2, we used foreign firms present in1998–2007 to calculate FDI-related variables. For-eign firms that entered China before 1998, as longas they were present in 1998 or after, were includedin the calculation. However, it is possible that someforeign firms entered China before 1998 and exitedbefore 1998. These foreign firms may have createdspillovers or competition effects on domestic firmsprior to their exit. In this sense, our data may be

left-censored. To address this issue, we re-esti-mated the models in Table 2 by focusing only ondomestic firms that were founded in 1998 or after.These domestic firms did not coexist with foreignfirms that exited prior to 1998, and thus were notaffected by those foreign firms. Using this subgroupof domestic firms thus enabled us to address theissue of left-censoring. The results are reported inTable 4, and they are highly consistent with thosereported in Table 2. Thus, we are confident that ourresults were not driven by the left-censoring issue.

Supplementary Analysis III: Effects of ForeignFirms in Different Geographic Locations

In Table 2, we used all foreign firms in an indus-try, in a year, to calculate FDI-related variables.This approach assumed that all foreign firms in theindustry were relevant, albeit their relevance wasweighted by their sizes, to the focal domestic firms.However, geographic proximity may make differ-ences: foreign firms located in the same region asthe focal domestic firms may matter more to thefocal domestic firms than those outside the region.To address this issue, we ran two supplementary

TABLE 4Firm-Fixed Effect Models of Production Function of Domestic Firms Founded in 1998 or Latera,b


Log K 0.340*** (0.003) 0.338*** (0.003) 0.338*** (0.003)Log L 0.333*** (0.003) 0.333*** (0.003) 0.333*** (0.003)PredictorsEntry tenure of foreign firms in an industry 0.008*** (7.42E-04) 0.008*** (7.41E-04)Export intensity of foreign firms –0.064*** (0.008) –0.065*** (0.008)Intangible asset intensity of foreign firms –0.321*** (0.036) –0.339*** (0.037)Rhythm of entry pattern of foreign firms 0.004*** (7.27E-04) 0.003*** (7.29E-04)InteractionsSquared term of entry tenure of foreign

firms in an industry�9.56E-04*** (1.76E-04)

Entry tenure of foreign firms in an industry �Export intensity of foreign firms

–0.011*** (0.002)

Entry tenure of foreign firms in an industry �Intangible asset intensity of foreign firms

–0.042** (0.015)

Entry tenure of foreign firms in an industry �Rhythm of entry pattern of foreign firms

0.003*** (3.74E-04)

Controls(All controls listed in Table 2 were included.)F value 1,119*** 1,070*** 1,022***Within R2 0.2773 0.2780 0.2783Total R2 0.9290 0.9291 0.9292

a n � 438,729 firm-year observations.b Estimated coefficients and associated robust standard errors (in parentheses) are reported.

* p � 0.05 (two-tailed tests)** p � 0.01

*** p � 0.001


analyses. In Model A of Table 5, FDI-related vari-ables were calculated using data on foreign firms inthe same province as the focal domestic firms only(i.e., within-region foreign firms). In Model B ofTable 5, FDI-related variables were calculated us-ing data on foreign firms outside the province of thefocal domestic firms (i.e., outside-region foreignfirms). Our hypotheses are generally supported bythe results of both analyses, except that the moder-ating effects of foreign firms’ intangible asset inten-sity (Hypothesis 4) and the rhythm of their entrypattern (Hypothesis 5) are not supported inModel A (within-region foreign firms).

Supplementary Analysis IV: Effects of ForeignFirms With Different Country Origins

Foreign firms’ country origins may also matter.Buckley et al. (2007) and Chang and Xu (2008)divided foreign firms in China into two groups:those from Hong Kong, Macau, and Taiwan (here-after, “HMT”) and those from other countries. Dueto these regions’ historical and social linkages withmainland China, foreign firms from HMT havesome access to local knowledge and local resourcesand thus have greater resource similarity with do-mestic firms than with foreign firms from othercountries (Chang & Xu, 2008). To check for anypotentially different effects of these two groups offoreign firms, in Model C of Table 5, FDI-relatedvariables were calculated using data on foreignfirms from HMT only, while, in Model D, FDI-related variables were calculated using data on for-eign firms from other countries. Results of thesetwo analyses consistently support the predictionsof Hypotheses 1–4, but not that of Hypothesis 5.

Supplementary Analysis V: AlternativeDefinitions of “Foreign Firms” and“Domestic Firms”

In earlier analyses, “foreign firms” referred tofirms with 100% foreign ownership and “domesticfirms” referred to firms with 100% domestic own-ership. We used alternative cut-offs to define do-mestic firms and foreign firms. Table 6 summarizesthe results. In Models A–C, domestic firms referredto those with domestic ownership greater than 95%(Model A), 80% (Model B), and 75% (Model C),respectively (foreign firms referred to those with100% foreign ownership). In Models D–F, foreignfirms referred to those with foreign ownershipgreater than 95% (Model D), 80% (Model E), and

75% (Model F), respectively (domestic firms re-ferred to those with 100% domestic ownership).Results of these models are highly consistent withone another as well as with those reported in Ta-ble 2. Thus, our results are robust to alternativedefinitions of “domestic firms” and “foreign firms.”

Supplementary Analysis VI: Using Longer Lags

In the analyses reported in Table 2, we used a1-year lag for predictors. For robustness checks, weused longer lags. The results are reported in Ta-ble 7, in which Models A–D used 2-, 3-, 4-, and5-year lags, respectively. The entry tenure of for-eign firms in an industry is positive and significantand its squared term is negative and significantacross these models, consistently supporting Hy-potheses 1 and 2. It is not surprising that the effectsize of the entry tenure of foreign firms is smaller inmodels with longer lags (b � 0.021, 0.007, 0.006,and 0.004 for 2-, 3-, 4-, and 5-year lags, respec-tively). Among the three moderating effects, themoderating effect of foreign firms’ export intensity(Hypothesis 3) is the most robust and that of foreignfirms’ entry pattern rhythm (Hypothesis 5) is theleast robust. Nevertheless, none of the moderatingeffects is supported in Model D, where a 5-year lagwas used.

Supplementary Analysis VII: Using RandomSub-Samples

Considering our large sample size, we randomlydrew 50, 20, and 10% of the domestic firms in ourdata and re-estimated our models. These analysesproduced results highly consistent with those re-ported in this paper. For the sake of space, theseadditional results are not reported here, but areavailable from the authors upon request.

DISCUSSION AND CONCLUSIONS

Contributions to the FDI Spillover Literature

Our study was motivated by the desire to betterunderstand how FDI spillovers may occur overtime in an emerging market. It contributes to theliterature by explicitly investigating the dynamicsof the spillover process. As noted earlier, priorresearch has primarily taken a snapshot approachto examine how FDI presence (in terms of the shareof FDI in an industry) affects domestic firms. Whileseveral studies have acknowledged that time is a


TA

BL

E5

Fir

m-F

ixed

Eff

ect

Mod

els

ofD

omes

tic

Fir

mP

rod

uct

ion

Fu

nct

ion

—E

ffec

tsof

For

eign

Fir

ms

byG

eogr

aph

icL

ocat

ion

san

dC

oun

try

Ori

gin

sa,b

For

eign

Fir

ms

byG

eogr

aph

icL

ocat

ion

sF

orei

gnF

irm

sby

Cou

ntr

yO

rigi

ns

Var

iabl

esM

odel

A(w

ith

inth

ere

gion

)M

odel

B(o

uts

ide

the

regi

on)

Mod

elC

(fro

mH

MT

b)

Mod

elD

(fro

mot

her

cou

ntr

ies)

Log

K0.

364*

**(0

.002

)0.

376*

**(0

.002

)0.

373*

**(0

.002

)0.

374*

**(0

.002

)L

ogL

0.36

9***

(0.0

03)

0.37

6***

(0.0

02)

0.37

8***

(0.0

02)

0.37

6***

(0.0

02)

Pre

dic

tors

En

try

ten

ure

offo

reig

nfi

rms

inan

ind

ust

ry0.

009*

**(4

.94E

-04)

0.00

8***

(4.0

7E-0

4)0.

014*

**(3

.31E

-04)

0.01

0***

(2.8

8E-0

4)E

xpor

tin

ten

sity

offo

reig

nfi

rms

–0.0

73**

*(0

.005

)–0

.029

***

(0.0

04)

�7.

03E

-04

(0.0

03)

–0.0

24**

*(0

.004

)In

tan

gibl

eas

set

inte

nsi

tyof

fore

ign

firm

s–0

.401

***

(0.0

30)

–0.0

31**

(0.0

11)

�0.

007

(0.0

08)

–0.1

21**

*(0

.017

)R

hyt

hm

ofen

try

pat

tern

offo

reig

nfi

rms

0.00

4***

(4.1

8E-0

4)0.

002*

**(3

.99E

-04)

�7.

00E

-04

(4.0

0E-0

4)1.

46E

-03*

**(3

.28E

-04)

Inte

ract

ion

sS

quar

edte

rmof

entr

yte

nu

reof

fore

ign

firm

sin

anin

du

stry

�7.

98E

-04*

**(1

.12E

-04)

�8.

89E

-04*

**(8

.80E

-05)

�8.

49E

-04*

**(6

.48E

-05)

�5.

69E

-04*

**(5

.23E

-05)

En

try

ten

ure

offo

reig

nfi

rms

inan

ind

ust

ry�

Exp

ort

inte

nsi

tyof

fore

ign

firm

s–0

.008

***

(1.5

3E-0

3)–0

.004

***

(1.2

2E-0

3)–0

.004

***

(8.9

5E-0

4)–0

.004

***

(8.5

6E-0

4)

En

try

ten

ure

offo

reig

nfi

rms

inan

ind

ust

ry�

Inta

ngi

ble

asse

tin

ten

sity

offo

reig

nfi

rms

�0.

005

(0.0

11)

–0.0

12**

(0.0

04)

–0.0

15**

*(0

.003

)–0

.037

***

(0.0

08)

En

try

ten

ure

offo

reig

nfi

rms

inan

ind

ust

ry�

Rh

yth

mof

entr

yp

atte

rnof

fore

ign

firm

s4.

86E

-05

(1.8

2E-0

4)4.

90E

-04*

*(1

.83E

-04)

1.57

E-0

4(1

.31E

-04)

�1.

64E

-04

(1.1

9E-0

4)

Con

trol

s(A

llco

ntr

ols

list

edin

Tab

le2

wer

ein

clu

ded

.)F

valu

e1,

460*

**2,

748*

**2,

707*

**2,

680*

**W

ith

inR

20.

2834

0.28

450.

2855

0.28

72T

otal

R2

0.93

080.

9151

0.91

480.

9135

N(f

irm

–yea

rob

serv

atio

ns)

638,

185

937,

013

917,

982

896,

967

aE

stim

ated

coef

fici

ents

and

asso

ciat

edro

bust

stan

dar

der

rors

(in

par

enth

eses

)ar

ere

por

ted

.b

HM

Tre

fers

toH

ong

Kon

g,M

acau

,an

dT

aiw

an.

*p

�.0

5(t

wo-

tail

edte

sts)

**p

�.0

1**

*p

�.0

01

TA

BL

E6

Fir

m-F

ixed

Eff

ect

Mod

els

ofD

omes

tic

Fir

mP

rod

uct

ion

Fu

nct

ion

—A

lter

nat

ive

Def

init

ion

sof

Dom

esti

cF

irm

san

dF

orei

gnF

irm

sa,b

,c

Dom

esti

cF

irm

s,in

Wh

ich

Dom

esti

cO

wn

ersh

ipis

..

.bF

orei

gnF

irm

s,in

Wh

ich

For

eign

Ow

ner

ship

is.

..c

>95

%>

80%

>75

%>

95%

>80

%>

75%

Var

iabl

esM

odel

AM

odel

BM

odel

CM

odel

DM

odel

EM

odel

F

Log

K0.

379*

**(0

.002

)0.

381*

**(0

.002

)0.

381*

**(0

.002

)0.

379*

**(0

.002

)0.

378*

**(0

.002

)0.

378*

**(0

.002

)L

ogL

0.38

2***

(0.0

03)

0.38

3***

(0.0

03)

0.38

4***

(0.0

03)

0.38

0***

(0.0

03)

0.38

1***

(0.0

03)

0.38

1***

(0.0

03)

Pre

dic

tors

En

try

ten

ure

offo

reig

nfi

rms

inan

ind

ust

ry0.

016*

**(4

.74E

-04)

0.01

6***

(4.7

2E-0

4)0.

016*

**(4

.70E

-04)

0.01

6***

(4.8

0E-0

4)0.

018*

**(5

.08E

-04)

0.01

9***

(5.0

5E-0

4)

Exp

ort

inte

nsi

tyof

fore

ign

firm

s–0

.050

***

(0.0

05)

–0.0

50**

*(0

.005

)–0

.050

***

(0.0

05)

–0.0

49**

*(0

.005

)–0

.056

***

(0.0

05)

–0.0

68**

*(0

.005

)

Inta

ngi

ble

asse

tin

ten

sity

offo

reig

nfi

rms

–0.0

29**

(0.0

11)

–0.0

30**

(0.0

11)

–0.0

31**

(0.0

11)

�0.

020

(0.0

11)

–0.0

76**

*(0

.013

)–0

.050

***

(0.0

14)

Rh

yth

mof

entr

yp

atte

rnof

fore

ign

firm

s0.

005*

**(4

.14E

-04)

0.00

5***

(4.1

2E-0

4)0.

005*

**(4

.11E

-04)

0.00

5***

(4.1

8E-0

4)0.

007*

**(4

.26E

-04)

0.00

7***

(3.9

9E-0

4)

Inte

ract

ion

sS

quar

edte

rmof

entr

yte

nu

reof

fore

ign

firm

sin

anin

du

stry

–0.0

02**

*(1

.16E

-04)

–0.0

02**

*(1

.15E

-04)

–0.0

02**

*(1

.15E

-04)

–0.0

02**

*(1

.19E

-04)

–0.0

02**

*(1

.25E

-04)

–0.0

03**

*(1

.25E

-04)

En

try

ten

ure

offo

reig

nfi

rms

inan

ind

ust

ry�

Exp

ort

inte

nsi

tyof

fore

ign

firm

s

–0.0

06**

*(1

.50E

-03)

–0.0

06**

*(1

.50E

-03)

–0.0

06**

*(1

.49E

-03)

–0.0

08**

*(1

.52E

-03)

–0.0

13**

*(1

.65E

-03)

–0.0

11**

*(1

.63E

-03)

En

try

ten

ure

offo

reig

nfi

rms

inan

ind

ust

ry�

Inta

ngi

ble

asse

tin

ten

sity

offo

reig

nfi

rms

–0.0

32**

*(0

.004

)–0

.033

***

(0.0

04)

–0.0

33**

*(0

.004

)–0

.035

***

(0.0

05)

–0.0

72**

*(0

.006

)–0

.080

***

(0.0

06)

En

try

ten

ure

offo

reig

nfi

rms

inan

ind

ust

ry�

Rh

yth

mof

entr

yp

atte

rnof

fore

ign

firm

s

7.09

E-0

4***

(2.1

8E-0

4)7.

36E

-04*

**(2

.18E

-04)

7.48

E-0

4***

(2.1

7E-0

4)6.

35E

-04*

*(2

.23E

-04)

0.00

1***

(2.4

4E-0

4)7.

97E

-04*

**(2

.29E

-04)

Con

trol

s(A

llco

ntr

ols

inT

able

2w

ere

incl

ud

ed.)

Fva

lue

1,94

1***

1,96

8***

1,98

2***

1,98

3***

1,98

7***

1,99

1***

Wit

hin

R2

0.26

650.

2681

0.26

850.

2656

0.26

580.

2660

Tot

alR

20.

9310

0.93

150.

9317

0.93

090.

9309

0.93

09N

(fir

m-y

ear

obse

rvat

ion

s)94

7,43

295

4,43

696

0,47

094

5,55

394

5,55

394

5,55

3

aE

stim

ated

coef

fici

ents

and

asso

ciat

edro

bust

stan

dar

der

rors

(in

par

enth

eses

)ar

ere

por

ted

.b

InM

odel

sA

–C,

fore

ign

firm

sre

ferr

edto

thos

ew

ith

100%

fore

ign

own

ersh

ip.

cIn

Mod

els

D–G

,d

omes

tic

firm

sre

ferr

edto

thos

ew

ith

100%

dom

esti

cow

ner

ship

.*

p�

.05

(tw

o-ta

iled

test

s)**

p�

.01

***

p�

.001

TA

BL

E7

Fir

m-F

ixed

Eff

ect

Mod

els

ofD

omes

tic

Fir

mP

rod

uct

ion

Fu

nct

ion

—L

onge

rL

agsa

Var

iabl

esM

odel

AM

odel

BM

odel

CM

odel

D

2-ye

arla

g3-

year

lag

4-ye

arla

g5-

year

lag

Log

K0.

383*

**(0

.003

)0.

383*

**(0

.003

)0.

396*

**(0

.004

)0.

412*

**(0

.004

)L

ogL

0.38

6***

(0.0

03)

0.42

2***

(0.0

04)

0.44

7***

(0.0

05)

0.46

2***

(0.0

06)

Pre

dic

tors

En

try

ten

ure

offo

reig

nfi

rms

inan

ind

ust

ry0.

021*

**(5

.46E

-04)

0.00

7***

(6.3

6E-0

4)0.

006*

**(8

.12E

-04)

0.00

4***

(0.0

01)

Exp

ort

inte

nsi

tyof

fore

ign

firm

s–0

.046

***

(0.0

05)

–0.0

13*

(0.0

06)

0.00

3(0

.007

)�

0.00

3(0

.009

)In

tan

gibl

eas

set

inte

nsi

tyof

fore

ign

firm

s�

0.00

2(0

.012

)0.

028

(0.0

16)

0.01

3(0

.021

)�

0.03

2(0

.036

)R

hyt

hm

ofen

try

pat

tern

offo

reig

nfi

rms

0.00

5***

(4.5

8E-0

4)4.

88E

-06

(5.2

9E-0

4)–0

.001

(6.2

7E-0

4)�

9.81

E-0

4(7

.48E

-04)

Inte

ract

ion

sS

quar

edte

rmof

entr

yte

nu

reof

fore

ign

firm

sin

anin

du

stry

–0.0

02**

*(1

.35E

-04)

�7.

82E

-04*

**(1

.45E

-04)

�8.

69E

-04*

**(1

.91E

-04)

–3.6

6E-0

4(2

.28E

-04)

En

try

ten

ure

offo

reig

nfi

rms

inan

ind

ust

ry�

Exp

ort

inte

nsi

tyof

fore

ign

firm

s–4

.55E

-03*

*(1

.75E

-03)

–0.0

07**

*(1

.90E

-03)

–0.0

11**

*(2

.34E

-03)

–0.0

01(2

.96E

-03)

En

try

ten

ure

offo

reig

nfi

rms

inan

ind

ust

ry�

Inta

ngi

ble

asse

tin

ten

sity

offo

reig

nfi

rms

–0.0

17**

*(4

.58E

-03)

�0.

004

(0.0

05)

–0.0

17**

(0.0

06)

0.00

2(0

.007

)

En

try

ten

ure

offo

reig

nfi

rms

inan

ind

ust

ry�

Rh

yth

mof

entr

yp

atte

rnof

fore

ign

firm

s�

1.70

E-0

4(2

.51E

-04)

5.73

E-0

4*(2

.85E

-04)

3.21

E-0

4(3

.45E

-04)

�2.

97E

-04

(4.0

2E-0

4)

Con

trol

s(A

llco

ntr

ols

inT

able

2w

ere

incl

ud

ed.)

Fva

lue

1,07

9***

1,91

9***

1,36

2***

1,19

6***

Wit

hin

R2

0.24

390.

3436

0.36

900.

4072

Tot

alR

20.

9397

0.90

880.

9059

0.90

30N

(Fir

m–y

ear

obse

rvat

ion

s)64

4,20

842

3,28

125

9,45

716

7,20

4

aE

stim

ated

coef

fici

ents

and

asso

ciat

edro

bust

stan

dar

der

rors

(in

par

enth

eses

)ar

ere

por

ted

.*

p�

0.05

(tw

o-ta

iled

test

s)**

p�

0.01

***

p�

0.00

1

relevant dimension in assessing FDI spillovers(e.g., De Backer & Sleuwaegen, 2003; Spencer,2008), ours is the first one that directly examinedthe role of time in FDI spillovers. Using a compre-hensive panel dataset of manufacturing firms inChina in 1998–2007, we found that the entry ten-ure of foreign firms in an industry had a positiverelationship with the productivity of individual do-mestic firms in the industry, and this positive rela-tionship became weaker as the entry tenure of theforeign firms continued to increase. These resultswere further validated by using domestic firms’survival as an alternative outcome variable. Ourfindings suggest that it is not just the presence offoreign firms but also the entry tenure of the foreignfirms that affects their spillovers to domestic firms.Our results also suggest that there may be a limit onthe extent to which FDI spillovers occur over timebecause obvious opportunities for domestic firmsto learn would gradually dry up.

We further drew upon the competitor imitationargument (Lippman & Rumelt, 1982; Reed & DeFil-lippi, 1990) to examine how the role of time in FDIspillovers may depend upon the imitability of theforeign firms. We found that the positive relation-ship between the entry tenure of foreign firms in anindustry and the productivity of individual domes-tic firms in the industry was stronger when theforeign firms had lower export intensity, lower in-tangible asset intensity, and had followed a morerhythmic entry pattern—attributes that are associ-ated with lower barriers to imitation. These resultsare consistent with our argument that the role oftime in FDI spillovers depends upon the height ofbarriers of imitation faced by the domestic firms:the lower the barriers, the greater the effect of timein FDI spillovers. While not hypothesized, wefound that the three moderating variables had directrelationships with the productivity of individual do-mestic firms—export intensity and intangible assetsintensity of foreign firms had significantly negativerelationships, and the rhythm of their entry patternhad a significantly positive relationship.

Our focus on the role of time in FDI spilloversdiffers from and yet complements some recent re-search efforts in this area. Chang and Xu (2008)separated FDI presence at the national level and atthe regional level. They found that FDI presence atthe national level increased the survival rate ofdomestic firms while FDI presence at the regionallevel reduced the survival rate of domestic firms.Zhang et al. (2010) focused on the diversity ofcountry origins of foreign firms in an industry and

found that it had a significantly positive relation-ship with the productivity of domestic firms in theindustry. These studies advanced the literature byexamining the (intra-temporal) heterogeneous na-ture of FDI. In contrast, our study dealt with theinter-temporal aspect of FDI spillovers. Ourfindings show that, as the entry tenure of foreignfirms in an industry increased, domestic firms inthe industry improved their productivity/increasedtheir survival rate (albeit at a diminishing rate), andthis effect varied across industries, dependingupon the imitability of the foreign firms. Our inter-temporal approach added an important new di-mension to FDI spillover research, and should en-courage further research in this direction.

Moreover, our focus on the contingent effects offoreign firms’ attributes also differs from and com-plements recent studies in this area. Recognizingthe importance of domestic firms as knowledgerecipients in the FDI spillover relationship, recentstudies have examined how attributes of domesticfirms—particularly, their motivation and ability tolearn—can affect the extent to which they can ben-efit from FDI spillovers (e.g., Sjöholm, 1999; Zhanget al., 2010). However, foreign firms, as the knowl-edge source firms in the FDI spillover relationship,have been largely treated as “black boxes” (Spen-cer, 2008). Our results suggest foreign firms’ attri-butes—that is, their export intensity, intangible as-set intensity, and the rhythm of their entrypattern—affected the speed with which FDI spill-overs occur. These attributes of foreign firms alsohad direct relationships with the productivity ofdomestic firms. Our study and other recent studiesin the area jointly demonstrate that we need toconsider the attributes of both foreign firms anddomestic firms in order to better understand howspillovers occur between them.

Practical Implications

Our findings have important implications for MNCmanagers, domestic managers, and policy makers inemerging markets. For MNC managers, our findingshelp them better understand how domestic firms inthe host country learn from them over time, and un-der what conditions they are more or less vulnerableto domestic imitation. As noted in a recent reportissued by the Economist Intelligence Unit (2011: 4):“It is often taken as fact that multinationals havesuperior technology and better brand management. . . .There are signs that all of these advantages arebeginning to erode in China.” Our findings suggest


that, while domestic firms are catching up overtime, the speed of catching up varies across indus-tries—the speed is slower in industries with higherlevels of imitation barriers.

For domestic managers in emerging markets, ourfindings send a clear message to them that benefitsof FDI spillovers take time to materialize, and thereis also a limit on the extent to which domestic firmscan learn from foreign firms over time. Figure 1illustrates that, as entry tenure of foreign firms inan industry reaches two to three standard devia-tions above its sample mean (about 10–12 years),the slope becomes close to being flat. These resultssuggest that 10–12 years is a reasonable time framefor FDI spillovers, after which domestic firms’ ad-ditional productivity gains through learning fromforeign firms would be limited. Our findings alsosuggest that domestic firms need to find the right“targets” for learning: foreign firms with local mar-ket focus and those with high tangible assets inten-sity are relatively easier to imitate.

Our results also provide suggestions for policymakers in emerging markets on how to manage theprocess of FDI inflow in order to maximize FDIspillovers. Bringing in FDI at a more rhythmic paceenables domestic firms to better observe and under-stand foreign entrants’ technologies, and, thus, tobenefit more from FDI spillovers. Also, govern-ments of emerging markets very often pressure for-eign firms to export because export growth isusually closely associated with the economic de-velopment of emerging markets. Our results dem-onstrate that foreign firms’ export may delay thespeed with which domestic firms learn from theforeign firms. Thus, policy makers need to take intoaccount the potential learning benefits of domesticfirms when making their FDI policies.

Limitations and Future Research Directions

This study has some limitations that offer fruitfulopportunities for future research in this area. First,we examined the role of time in FDI spillovers byfocusing on the entry tenure of foreign firms in anindustry. Our findings provide an overall picture ofhow foreign firms’ spillovers to domestic firms oc-cur over time. Future studies may examine morespecific dimensions of time; for example, timeelapsed since some milestone events related to FDI.One possible milestone event is a significant in-vestment in an industry made by a major MNC.Such an investment signals the potential of theindustry in the host country, which will attract

attention from both foreign firms and domesticfirms to this sector. Also, considering the impor-tance of local suppliers in FDI spillovers, whenforeign firms started to purchase from local suppli-ers (or when their overseas suppliers were relo-cated to the host country) may be another mile-stone event.

Second, we examined three attributes of foreignfirms that may affect barriers to imitation; future re-search may examine other sources of barriers to imi-tation. One possible direction is to look at foreignfirms’ strategic alliances with domestic firms. Foreignfirms’ knowledge transferred to their domestic part-ners may be further passed to domestic non-partnerfirms, with the domestic partners acting as mediatingchannels (Spencer, 2008). Thus, to the extent thatforeign firms in an industry are prone to formingstrategic alliances with domestic firms, the barriers toimitation may be much reduced.

Third, we focused on FDI spillovers in China; thequestion of whether our findings are generalizableto other emerging markets remains unanswered.We believe that the core ideas that FDI spilloverstend to occur over time (at a diminishing rate) andthat the attributes of foreign firms may affect therole of time are applicable to other emerging mar-kets. Emerging markets, however, are differentalong many dimensions, including the maturity ofeconomic growth, institutional stability, and thelevel of protection of property rights and contractenforcement (Hitt et al., 2005). Compared to otheremerging markets, such as Romania, Kenya, andEcuador, China has some unique attributes, in-cluding strong industrial policy, governmentownership, and enormous internal markets. Yet,how to attract FDI and to facilitate FDI spilloversrepresent major policy challenges for manyemerging markets. Future research should repli-cate our study with firms in other emerging mar-kets. Such research effort cannot only add to ourknowledge on FDI spillovers, but also providevaluable practical suggestions on how emergingmarket countries can develop their economies byusing FDI.

In conclusion, to the best of our knowledge, this isthe first empirical study that explicitly examines howforeign firms’ spillovers to domestic firms may occurover time in an emerging market and how foreignfirms’ attributes may affect the effect of time in thespillover process. Our findings can contribute to abetter understanding of the dynamics of FDI spill-overs and should stimulate future research effort tofocus on this emerging and interesting research topic.


REFERENCES

Aiken, L. S., & West, S. G. 1991. Multiple regression:Testing an interpreting interactions. NewburyPark, CA: Sage.

Aitken, B. J., & Harrison, A. E. 1999. Do domestic firmsbenefit from direct foreign investment? Evidencefrom Venezuela. American Economic Review, 89:605–618.

Argyris, C., & Schön, D. 1978. Organizational learning:A theory of action perspective. Reading, MA: Ad-dison-Wesley.

Barth, M. E., Kasznik, R., & McNichols, M. F. 2001. An-alyst coverage and intangible assets. Journal of Ac-counting Research, 39: 1–34.

Belderbos, R., Capannelli, G., & Fukao, K. 2001. Back-ward vertical linkages of foreign manufacturing af-filiates: Evidence from Japanese multinationals.World Development, 29: 189–208.

Blomström, M., & Kokko, A. 1998. Multinational corpo-rations and spillovers. Journal of Economic Sur-veys, 12: 247–277.

Brown, S. L., & Eisenhardt, K. M. 1997. The art of contin-uous change: Linking complexity theory and time-paced evolution in relentlessly shifting organizations.Administrative Science Quarterly, 42: 1–34.

Buckley, P. J., Clegg, J., & Wang, C. 2007. Is the relation-ship between inward FDI and spillover effects linear?An empirical examination of the case of China. Jour-nal of International Business Studies, 38: 447–459.

Carroll, G. R., & Delacroix, J. 1982. Organizational mor-tality in the newspaper industries of Argentina andIreland: An ecological approach. AdministrativeScience Quarterly, 27: 169–198.

Caves, R. E. 1996. Multinational enterprise and eco-nomic analysis (2nd ed.). Cambridge: CambridgeUniversity Press.

Chang, S. J., & Xu, D. 2008. Spillovers and competitionamong foreign and local firms in China. StrategicManagement Journal, 29: 495–518.

Chatterjee, A., & Hambrick, D. C. 2007. It’s all about me:Narcissistic chief executive officers and their effectson company strategy and performance. Administra-tive Science Quarterly, 52: 351–386.

Chung, W., Mitchell, W., & Yeung, B. 2003. Foreign di-rect investment and host country productivity: TheAmerican automotive component industry in the1980s. Journal of International Business Studies,34: 199–218.

Cohen, W. M., & Levinthal, D. A. 1990. Absorptive capacity:A new perspective on learning and innovation. Ad-ministrative Science Quarterly, 35: 128–152.

Cyert, R. M., & March, J. G. 1963. A behavioral theory ofthe firm. Englewood Cliffs, NJ: Prentice Hall.

De Backer, K., & Sleuwaegen, L. 2003. Does foreign directinvestment crowd out domestic entrepreneurship?Review of Industrial Organization, 22: 67–102.

Dierickx, I., & Cool, K. 1989. Asset stock accumulationand sustainability of competitive advantage. Man-agement Science, 35: 1504–1511.

Economist Intelligence Unit. 2011. Multinational com-panies and China: What future? EIU report, 15 De-cemeber 2011, London.

Eden, L. 2009. Letter from the editor-in-chief: FDI spill-overs and linkages. Journal of International Busi-ness Studies, 40: 1065–1069.

Feinberg, S. E., & Gupta, A. K. 2009. MNC subsidiariesand country risk: Internalization as a safeguard againstweak external institutions. Academy of ManagementJournal, 52: 381–399.

Feinberg, S. E., & Majumdar, S. K. 2001. Technologyspillovers from foreign direct investment in the In-dian pharmaceutical industry. Journal of Interna-tional Business Studies, 32: 421–437.

Freeman, J., Carroll, G. R., & Hannan, M. T. 1983. Theliability of newness: Age dependence in organiza-tional death rates. American Sociological Review,48: 692–710.

Görg, H., & Strobl, E. 2001. Multinational companies andproductivity spillovers: A meta-analysis. EconomicJournal, 111: 723–739.

Hitt, M. A., Li, H., & Worthington, W. J. 2005. Emergingmarkets as learning laboratories: Learning behaviorsof local firms and foreign entrants in different insti-tutional contexts. Management and OrganizationReview, 1: 353–380.

Inkpen, A. C., & Tsang, E. W. K. 2005. Social capital,networks, and knowledge transfer. Academy ofManagement Review, 30: 146–165.

Javorcik, B. S. 2004. Does foreign direct investment in-crease the productivity of domestic firms? In searchof spillovers through backward linkages. AmericanEconomic Review, 94: 605–627.

Kogut, B., & Zander, U. 1992. Knowledge of the firm,combinative capabilities, and the replication of tech-nology. Organization Science, 3: 383–397.

Kosová, R. 2010. Do foreign-invested firms crowd outdomestic firms? Evidence from the Czech Republic.Review of Economics and Statistics, 92: 861–881.

Lippman, S. A., & Rumelt, R. P. 1982. Uncertain imita-bility: An analysis of interfirm differences in effi-ciency under competition. Bell Journal of Econom-ics, 13: 418–438.

Luo, Y., & Park, S. H. 2001. Strategic alignment and


performance of market-seeking MNCs in China.Strategic Management Journal, 22: 141–155.

Markusen, J. R. 1995. The boundaries of multinationalenterprises and the theory of international trade.Journal of Economic Perspectives, 9: 169–189.

Meyer, K. E., & Sinani, E. 2009. When and where doesforeign direct investment generate positive spill-overs? A meta-analysis. Journal of InternationalBusiness Studies, 40: 1075–1094.

Nelson, R., & Winter, S. 1982. An evolutionary theory ofeconomic change. Cambridge, MA: Belknap Press.

Osland, G., & Björkman, I. 1998. MNC-host governmentinteraction: Government pressures on MNCs inChina. European Management Journal, 16: 91–100.

Porter, M. E. 1985. Competitive advantage: Creatingand sustaining superior performance. New York:Free Press.

Reed, R., & DeFillippi, R. J. 1990. Causal ambiguity, barriersto imitation, and sustainable competitive advantage.Academy of Management Review, 15: 88–102.

Simon, H. A. 1991. Bounded rationality and organiza-tional learning. Organization Science, 2: 125–134.

Sjöholm, F. 1999. Technology gap, competition and spill-overs from direct foreign investment: Evidence fromestablishment data. Journal of Development Stud-ies, 36: 53–73.

Spencer, J. W. 2008. The impact of multinational enter-prise strategy on indigenous enterprises: Horizontalspillovers and crowding out in developing countries.Academy of Management Review, 33: 341–361.

Steensma, H. K., Tihanyi, L., Lyles, M. A., & Dhananraj,C. 2005. The evolving value of foreign partnershipsin transitioning economies. Academy of Manage-ment Journal, 48: 213–235.

Teece, D. J. 1998. Capturing value from knowledge assets:The new economy, markets for know-how, and in-tangible assets. California Management Review,40: 55–79.

Tian, X. 2007. Accounting for sources of FDI technologyspillovers: Evidence from China. Journal of Interna-tional Business Studies, 38: 147–159.

Vermeulen, F., & Barkema, H. 2002. Pace, rhythm, andscope: Process dependence in building a profitablemultinational corporation. Strategic ManagementJournal, 23: 637–654.

Winter, S. G. 2000. The satisficing principle in capabilitylearning. Strategic Management Journal, 21: 981–996.

Wooldridge, J. M. 2002. Econometric analysis of crosssection and panel data. Cambridge, MA: MIT Press.

Yan, A., & Gray, B. 1994. Bargaining power, managementcontrol, and performance in United States–Chinajoint ventures: A comparative case. Academy ofManagement Journal, 37: 1478–1517.

Zaheer, S., & Mosakowski, E. 1997. The dynamics of theliability of foreignness: A global study of survival infinancial services. Strategic Management Journal,18: 439–463.

Zhang, Y., Li, H., Hitt, M., & Cui, G. 2007. R&D intensityand international joint venture performance in anemerging market: Moderating effects of market focusand ownership structure. Journal of InternationalBusiness Studies, 38: 944–960.

Zhang, Y., Li, H., Li, Y., & Zhou, L. 2010. FDI spillovers inan emerging market: The role of foreign firms’ countryorigin diversity and domestic firms’ absorptive capac-ity. Strategic Management Journal, 31: 969–989.

Zhao, M. 2006. Conducting R&D in countries with weakintellectual property rights protection. ManagementScience, 52: 1185–1199.

Yan “Anthea” Zhang ([email protected]) is a professor ofstrategic management at the Jesse H. Jones GraduateSchool of Business, Rice University. She received herPh.D. from the Marshall School of Business, Universityof Southern California. Her current research interestsinclude CEO succession/dismissal, foreign direct invest-ment into and from emerging markets, and technologicalentrepreneurship in emerging markets.

Yu Li ([email protected]) is an associate professor of busi-ness strategy and international business at the BusinessSchool of the University of International Business and Eco-nomics in Beijing. She received her Ph.D. from GuanghuaSchool of Management, Peking University. Her current re-search interests include FDI spillover, the role of cluster infirm strategy and competitive advantage, and the interna-tionalization of firms from emerging markets.

Haiyang Li ([email protected]) is a professor of strategicmanagement and innovation at the Jesse H. Jones Grad-uate School of Business, Rice University. He received hisPh.D. from City University of Hong Kong. His currentresearch interests focus on technology entrepreneurshipand innovation, strategic alliances, and multinationalfirms’ innovation in emerging markets, as well as thegrowth of technology clusters in China.


Note: We created this illustration based upon Vermeulen and Barkema’s (2002: 642) Figure 1, which compared rhythmicand irregular patterns of firms’ internationalization processes.

APPENDIX ARhythmic vs. Irregular Patterns of Foreign Firms’ Entries

Ch

ange

in

th

e n

um

ber

of f

orei

gn fi

rms

in a

n

ind

ust

ry b

y ye

ar

Time Time

Rhythmic Pattern Irregular Pattern

Tot

al n

um

ber

of

fore

ign

firm

s in

an

in

du

stry

by

year


Copyright of Academy of Management Journal is the property of Academy of Managementand its content may not be copied or emailed to multiple sites or posted to a listserv withoutthe copyright holder's express written permission. However, users may print, download, oremail articles for individual use.

Documents

FDI SPILLOVERS OVER TIME IN AN EMERGING MARKET: THE …haiyang/AMJ FDI spillover 2014.pdfTHE ROLES OF ENTRY TENURE AND BARRIERS TO IMITATION YAN ANTHEA ZHANG Rice University ... spillovers